CN101688819A

CN101688819A - Be used for determining the method and the facility of Differential Group Delay and polarization mode dispersion

Info

Publication number: CN101688819A
Application number: CN200880018323.6A
Authority: CN
Inventors: N·西尔; 陈洪新
Original assignee: Exfo Electro Optical Engineering Inc
Current assignee: Exfo Electro Optical Engineering Inc
Priority date: 2007-03-28
Filing date: 2008-03-28
Publication date: 2010-03-31
Anticipated expiration: 2028-03-28
Also published as: WO2008116314A1; EP2140243A1; CN101688819B

Abstract

A kind of method that is used to measure at least one polarization-correlation properties of light path (FUT) is provided, this method is used near-end or near the light input element of near-end that is connected to this light path, near and the near-end or the far-end that are connected to this light path, or the light output unit near-end or the far-end.This light input element will have the light beam of the polarization at least in part of controlled polarization state (I-SOP) and go into this FUT.This output light unit extracts corresponding light from this FUT, analyze and detect the light that is extracted corresponding at least one transmission axle (A-SOP), and the corresponding electric signal of processing, with the transmission coherent light power under each wavelength of the light in each group that obtains at least two group wavelength, the wherein low described wavelength (λ in each described set of wavelengths _L) and high described wavelength (λ _U) closely alternate.Then calculate at least one poor corresponding in the power parameter that records of each wavelength of described wavelength centering for each batch total in described at least two groups, power parameter that this records and described analyze and the power of detected light is proportional subsequently define the set of at least two power parameter differences that record thus; Calculate the mean square value of the set of described difference; And calculate at least one and polarization-relevant FUT characteristic as at least one predefined function of described mean square value, described predefined function depends on that the little optical frequency between the described wavelength is poor, and described wavelength is corresponding to each the right wavelength in described at least two pairs of tight alternate wavelength.

Description

Be used for determining the method and the facility of Differential Group Delay and polarization mode dispersion

The cross reference of pertinent literature

The application requires the U.S. Provisional Patent Application No.60/907 of submission on March 28th, 2007, and the U.S. that on March 28th, 313 and 2007 submitted to partly continues patented claim No.11/727,759 right of priority.The full content of these two parts of patented claims will be included the present patent application text by reference in.

Technical field

The present invention relates to be used to measure the method and the facility of the polarization dependency characteristic of light path, and the present invention is particularly useful for measuring Differential Group Delay (DGD) or root mean square DGD specified wavelength interval in or the average DGD of a light path in the certain wave strong point, described light path mainly comprises optical waveguide, such as optical fiber link.When the specified wavelength interval was enough wide, root mean square DGD or average DGD measurement result extremely were worth near the polarization mode dispersion (PMD) of this light path.

Background technology

Orthogonal polarization modes at the optical fiber that is used for optical communication system has different group delays, i.e. the Differential Group Delay known to (DGD).This Differential Group Delay will cause polarization mode dispersion (PMD) phenomenon, i.e. pulse strenching behind the spread fiber.Under the situation that relates to the long optical fibers link, overall PMD can be enough to cause the increase of (scale-of-two) bit error rate, has so just limited transfer rate or maximum transmitted path.This especially is a problem under higher bit rate.Therefore, ideal situation is the PMD value that can obtain optical fiber.If the actual PMD value of known communication link just can accurately be estimated (scale-of-two) bit error rate or outage probability (failed probability in every period preset time of promptly communicating by letter), or Power penalty (needing the quantity of power that send into again for keeping the bit error rate identical when not having PMD) more.As the variable or the amount that characterize described PMD phenomenon, the PMD value of device is defined as root mean square (rms) value or the mean value of DGD.The DGD of a given device is a variable along with wavelength and time random variation.(, hereinafter, when suitable rms DGD or average DGD, will use " all DGD " sometimes for the purpose of simplifying.)

According to application, usually also need to measure equal DGD and the interior equal DGD of wide range of wavelengths in the DGD at a setted wavelength place, the narrow range of wavelengths.Yet, under many situations, can not measure the DGD at setted wavelength place or the equal DGD in the wide range of wavelengths, thereby can not realize by the measurement that given time carried out the reliable of PMD determined.

Below that's how things stand, for example when " PMD " in the narrow band communication channel measuring an optical fiber link, such as when the DWDM channel that dedicated bandwidth is arranged of the use about 70GHz of wavelength available width (corresponding to 100GHz DWDM channel separation) or about 35GHz (corresponding to the 50GHz channel separation) is measured.

" in the band " DGD measurement or equal DGD for given small wavelength interval in the channel measure the communication network provider that uses the DWDM network particularly important.It is a plurality of than on " secretly " channel on the telecommunication optical fiber link of low bit rate channel (for example 10Gbps) for example may to need to add one or more high bit rate channels (for example 40Gbps) to an activity delivered.In view of the harsh PMD tolerance limit under the higher bit rate, usually be necessary to characterize optical fiber link, or measure at least will the actual dark channel that uses, judge at the suitability that fully transmits aspect this high bit rate communication, and this measure characterize must the while do not interrupt those activities than low bit rate channel.

Ideally, should in a long-time section, carry out duplicate measurements off and on to the sign of single narrow channel because concerning single narrow channel, almost have only seldom or do not have possibility that the DGD that records is asked average on wavelength.If purpose is the PMD of measuring optical fiber link self, although then DWDM multiplexer/demultiplexer is attached to this optical fiber link, it is desirable to, in the dark channel that can obtain as much as possible, carry out in the band and measure.

Be used for measuring PMD method (end-to-end) of " broadband " optical fiber link and the method for the DGD in the narrow band channel on the measuring optical fiber at known in the art several.

The phase-shift method of being told about among the Jones (US4,750,833[4]) can be used to measure PMD.Of people such as Williams (Proceedings SOFM, Boulder CO, 1998, pp.23-26[5]), it also can be used for measuring the DGD of narrow band channel.Yet described method measurement itself is very slow, because it must make the institute's phase shift difference of surveying maximization by regulating Polarization Controller, so is unwell to wherein that optical fiber has relative to the fast outside line equipment that moves application.

" the impulse time delay method " that PMD measures can be measured the DGD that their corresponding light pulse differences time of arrival of exporting in the main attitude are measured the setted wavelength place by the fast polarization mode and the slow polarization mode that make short optical pulse enter optical fiber, but this method requires to use high speed electronics.Use polarization-interference short optical pulse, described based on detection such as people such as Noe (J.Lightwave Technology, Vol.20 (2), 2002, pp.229-235[6]) at time of arrival of this polarization-interference short optical pulse, can measure or estimate PMD.Yet this technology not only requires the high-velocity electrons detection system, and requires the light of fast modulation to be used for this measurement.

As people (J.Lightwave Technology such as people (US7256876) such as Yao (US 2005/020175 A1[7]) or Boroditsky and Wang, Vol.24 (11), 2006, pp.4120-4126[8]) the interior measurement facility of monitoring of band described, that use telecommunication real-time Communication for Power flow, allow directly to determine the PMD cost (promptly, for particular scene communication, for compensating extra system's surplus that the PMD impairment needs).Yet they can not determine the interior DGD of band or " PMD " value of this link.In fact, the interior monitoring method of these bands is for existing DOP or SOP monitoring under the high bit rate carrier signal situation to have advantage.(US 7.203 for people such as Waarts, 428, April 10,2007[9]) described and under tunable laser source, used heterodyne to detect the method that (heterodynedetection) estimates PMD, wherein combine with optical signalling from this link from the signal of local oscillator (being the tunable laser source), then analyze the amplitude of beat and the phase place of two orthogonal polarization modes states simultaneously, to obtain a SOP.Like this, average by a plurality of SOP can be estimated " PMD ".Yet same is that this measurement may only provide DOP or SOP information.The high-coherence light source that the method not only needs to add is used for this detection, and needs high-speed electronic components.

By people (J.Lightwave Technology such as use Wielandy, Vol.22 (3), 2004, pp.784-793[10]) described nonlinear detection techniques, can avoid the use of high-speed electronic components, but the application of nonlinear detection techniques will make the design of this apparatus complicated.

It should be noted that above-mentioned DOP or SOP measuring technique also may be subjected to amplified spontaneous emission (ASE), nonlinear fiber etc. and must influence (N.Kikuchi, Journal of Lightwave Technology, Vol.19 (4), 2001, pp.480-486[11]).Its susceptibility to ASE etc. is a major issue, because most long optical fibers link uses optical amplifier probably---and EDFA (Erbium-Doped Fiber Amplifier (EDFA)) or Raman optical amplifier.In addition, use SOP or the measurable DGD of DOP analytic approach interval very limited.

People such as C.D.Poole (J.Lightwave Technology, Vol.12 (6), 1994, pp.917-929[1]) described fixed analyzer (fixed analyzer) (or ground of equal value, length scanning) method is one of method that at first is applied to the PMD measurement, but uses big relatively range of wavelengths.When using small wavelength interval measure PMD, this technology all provides limited degree of accuracy.In addition, this method may not provide the DGD that depends on wavelength information.Therefore, this method also is not suitable for being used to measure narrow band channel.

Cyr J.Lightwave Technology, Vol.22 (3), 2004, pp.794-805 and US7,227,645[2,3] described general interferometric method (latter and the present invention are jointly owned), provide PMD measurement (corresponding to the spectrum width of broad band source) accurately, but this method can not provide the DGD that changes with wavelength, and not be to be suitable for very much narrow band channel.

Like this, DGD in current retrievable potentially, the arrowband separate channel that is suitable for measuring dwdm system or DGD or the PMD measuring technique of PMD, itself is very expensive, unreliable, have between limited dynamic area, may cause the instability in the quick gainequalizer, described balanced device usually appears in Reconfigurable Optical Add/drop Multiplexer (ROADM) and the optical amplifier.Like this, they are difficult to as feasible commercial instrument.

In view of the above, need a kind of new improved method so that can to in the DGD value carry out reliable measurements, cost appropriateness and have the measurement and the monitoring of high accuracy.Depend on application, the embodiment of the method should be able to respond the needs of " appropriate speed " monitoring (about 1 second of renewal speed) or " at a high speed " monitoring (about 1 millisecond of renewal speed).

In addition because the cause of convenience and operation cost aspect, when characterizing optical fiber, it is desirable to sometimes can be only from the overall PMD value of an end measuring optical fiber.But current, most methods of having developed that are used for on-the-spot PMD measurement mostly are " both-ends ", and promptly at one end (near-end) uses the specific polarization source, at the other end (far-end) operational analysis equipment [1,3].A kind of reliable and practical " single-ended " mensuration is not placed on far-end because do not need with specific light polarization source or other equipments being favourable aspect technician's tour and the logistics cooking.Also incite somebody to action/can be it is desirable to, can use identical technology or instrument to carry out single-ended or double-end measurement.

Be known that and use so-called single-ended PMD measuring technique,, come total PMD[12-14 of measuring optical fiber, 17] by only inserting FUT one end.Basically, the simplest single-ended PMD measurement mechanism comprises a tunable sharp CW light device [12,17] or tunable pulsed laser device [14]---between its output terminal and FUT, have a Polarization Controller (or polarization state generator) or the polarization polarizer; And has an analyzer, to analyze corresponding back reflective light.Usually,, or be admitted to this FUT from the pulsed light of this tunable pulsed laser device from the CW light of this tunable CW laser instrument, and analyzed from the back reflective light of the local reflection (such as Fresnel reflection) of this FUT far-end, to obtain total PMD value of this FUT.

Though single-ended PMD measures notion and method had before had progress, they are difficult to be developed as the feasible single-ended PMD that is used for and measure commercial instrument.Difficulty is the test and sensing device based on this conception of species, not too reliable, otherwise very expensive, or have the very long test duration, require optical fiber highly stable in long-time (promptly do not have robustness), perhaps only have between very limited dynamic area.

For example, for the single-ended PMD measuring technique of majority [12-16], tested optical fiber (FUT) should not move during measuring.This also is that so any optical fiber moves all will influence extreme value (being maximum value and minimal value), cause and estimate the PMD value mistakenly to conventional fixed analyzer method [13,15].For the single-ended scheme of fixed analyzer method, also may cause the mistake of DGD (or PMD) is estimated from any variable power of FUT back reflective light.Unfortunately in the measurement of reality, can not in the whole time period of measuring all data, guarantee this stability of FUT, especially under the situation of the DGD/PMD that measures mounted optical fiber.

And, list of references [13,15] the fixed analyzer method described in not only has strict demand to fibre movement, and also has a main latent defect aspect the measurement reliability, because this method only uses a detecting device to come measuring optical fiber definitely to lose (rather than standardized luminous power or transmission), and do not consider other latencies, such as fiber spectrum decay, the spectral losses of associated components that is used for an instrument or the gain that depends on wavelength of this detecting device.For example,, then will cause the error or the uncertainty of measurement result, especially for for the optical fiber with significant spectrum change (with respect to for the wavelength), as viewed in more outmoded optical cable if do not consider the spectrum decay of optical fiber.

In addition, no matter use the CW light source at those---be wideband light source or tunable laser [12,13,17] among---the known technology, measurement result all may be unreliable, because back reflective light may comprise from the remarkable contribution of rayleigh backscattering and from the local reflection of any falseness of the connector that is not positioned at this FUT far-end etc.Rayleigh contribution phenomenal growth along with the growth of fiber lengths, and reduce along with fiber lengths from the intensity of reflected light of local reflection (such as the Fresnel reflection at the FUT far-end), therefore make this CW light source method impracticable to the FUT length of several kms used in most telecommunications application.

Thereby, though the known technology that meets above-mentioned requirements may allow DGD/PMD is reasonably successfully measured at present, still be not enough to independently surveying instrument at present on their range of application and the performance as viable commercial.

Therefore, discussed, for example discuss in list of references [12-17], known technology and instrument can not be suitable for development of robust, reliable and low this commercial efficiently single-ended PMD test and sensing device easily.Measure total or overall PMD value exactly for an only end, be reported in available technology and concept in the works, current and have aforesaid limitation from optical fiber link.

In addition, as the U.S. Patent No. of owning together 6,724, explained in 469 (Leblanc) [18], in optical communication system, for specific long optical fibers, one or more short fiber sections of this optical fiber link can cause unacceptable overall polarization mode dispersion (PMD) level.For example wish to improve bit rate on the mounted optical fiber link in Internet service provider, as bring up under the situation of 40Gb/s, importantly can obtain distributed PMD measures, promptly obtain the PMD information that changes with respect to fiber distance, and locate one or more badly serious fiber segments so that it/they can be replaced, rather than change whole optical fiber.

In view of the above, Leblanc discloses the method that a kind of PMD of measurement distributes, and this method is used Polarization-Sensitive OTDR, discerning high or low PMD fiber segment, but does not provide the real quantitative PMD value of this FUT.Therefore, because " qualitative " person's character of himself, the technology of Leblanc is not suitable for being developed to the commercial single-ended overall PMD testing tool of the total PMD value that can measure whole optical fiber link on the whole.

Be known that and use so-called Polarization-Sensitive optical time domain reflectometer (POTDR; Generally be also referred to as " polarization optics domain reflectometer ") attempt to locate such " bad " fiber segment.Basically, POTDR is the responsive optical time domain reflectometer (OTDR) of a kind of polarization state to the back reflective signal (SOP).And conventional OTDR only measures the back reflective light intensity, to determine the variation of the decay of the length of---for example mounted optical fiber---along light path, and in order to monitor the polarization dependency characteristic of this transmission path, POTDR utilizes back reflective light also to show this fact of polarization dependence.Like this, the simplest POTDR comprises an OTDR, has a polarizer between its output terminal and this tested optical fiber (FUT); And between its photodetector and this FUT, comprising a polaroid analyze on the return path.Though (it will be appreciated that typical optical transmission path mainly comprises optical fiber, usually has other members, such as coupling mechanism, connector or the like in this path.Yet for convenience, other such members will be omitted, however should be understood that with in this article term " FUT " according to linguistic context will contain optical fiber and total transmission path the two.)

Usually, this POTDR can be divided into two classes or two kinds.First kind of POTDR is disclosed in document [19-24].

First kind of POTDR measures basically with this local birefringence that changes along the distance z of optical fiber (1/ claps length), or in other words, distributed birefringence.With reference to a simple and well-known delay wave plate example, birefringence is the phase delay (phase differential) of the per unit length between " slowly " axle and " soon " axle.In other words, delay is that wave plate thickness is multiply by in birefringence.Although general misunderstanding like this, this is not that PMD measures.At first, in the picture of simplifying, DGD (z) is the derivation thing of overall delay that extends to the fiber segment of z from 0, and change along with optical frequency (wavelength), secondly, long optical fibers shows as the cascade of a large amount of unit " wave plate ", to each unit wave plate, fast axle and the orientation of slow axis and the delay of per unit length change randomly along with distance z.

In view of the above, DGD (z) be to all drop on the upstream show along with the variation of z at random birefringence and the result of the complicated integration of the random orientation of this birefringence axis, and birefringence is the delay at some given position per unit length.In view of the above, as mentioned above, the DGD that limits in order to obtain, thing---it changes along with optical frequency must to use this integration derivation.

Therefore, the overall limitation of this first kind of technology is that they are not provided at the PMD on optical fiber distance along the line all direct, reliable, effective and quantitative under all scenario and measure.Opposite is that they measure local birefringence (or clapping long) and/or one or more correlation parameter, and release PMD from them---mainly based on supposition to this birefringent optic fibre characteristic and concrete pattern.For example, they suppose usually PMD and birefringent local value with and and so-called coupling length (or perturbation length) between a kind of relation, even this relation is effectively on average meaning, but also may not be exactly effective on local.

An example is that the strictness of this technology supposition optical fiber shows " linearity " birefringence.If there is circular birefringence really, because pass the roundtrip attribute (OTDR technology) of optical fiber, this circular birefringence also can " be missed " or be not in sight so.It should be noted that the correct measurement to modern " spin fiber " has required their performance is made supposition, is unacceptable to commercial apparatus therefore.

Second example is, even birefringence and other parameters must all measured on the whole length exactly---and in those sections of satisfied model of being supposed of optical fiber local characteristic and condition; Otherwise even if---it is the integration on some long length---said qualitatively, still can be estimated mistakenly to a great extent the PMD of these sections of being derived.In practice, though can measuring birefringence quantitatively, they (see people such as F.Corsa [19], on seeing), or statistics ground screening high birefringence section people [23] such as (, on seeing) Chen, or quantitatively the reaching relatively of PMD that obtains short section estimated, prerequisite be the accident of having approved frequent appearance (Leblanc[18], Huttner[22], on seeing), so this first kind of POTDR technology can not be reliably and measured in PMD, the especially scene PMD unknown, that mix the optical fiber of installing quantitatively.In addition, they can not, even approx, derive long length of fibre---such as the long optical fiber of 10 kms---overall PMD.

The claimed a kind of disclosed technology of people such as Fayolle [24] (on seeing), it is " real quantitative, at least in given polarization mode dispersion interval ".Yet this technology also runs into the essence limitation with this type association, as mentioned above.In fact, have two SOP (45 ° separately) of two tracks variation by utilization, possible output surmounts similar first kind of POTDR (for example, [23] of people such as Chen, the VOS of the document is identical with [24] track variation of people such as Fayolle basically) appropriateness improve, perhaps be improved as Doubly, this can not cause the last real quantitative PMD with acceptable accuracy of this FUT distance along the line to measure.One of this commercial measurement well-known with clap long (birefringence) relevant or be correlated with but do not represent the parameter of PMD coefficient.In fact, even the analog result in the disclosed instructions of people such as Fayolle has also shown 200% uncertain surplus.

It is desirable to, can obtain direct, quantitative PMD measures, promptly measure at PMD, measuring along each place of a series of positions of its length and as carrying out classical end-to-end PMD as this optical fiber termination along the cumulative actual at the discrete location place of optical fiber.This is because determine that the parameter of pulse strenching is PMD, rather than birefringence.If the actual PMD value of known communication link, so just can determine the scale-of-two bit error rate or outage probability (communication failed probability in a period of time) exactly, or Power penalty (for keeping the bit error rate identical when not having PMD, the quantity of power that send into of need) more more.

(in this manual, term " accumulation PMD " is used to be different from traditionally the overall PMD by end-to-end measurement.Because PMD is not the local amount, PMD (z) is PMD from 0 to z integration, and it is similar to cumulative probability, rather than similar to probability distribution.Certainly, when distance z equaled the overall length of this FUT, accumulation PMD equaled overall PMD.)

Second kind of known POTDR is exclusively used in PMD and measures.This type do not run into the essence limitation of above-mentioned first kind of POTDR, therefore shows the remarkable improvement that surmounts first kind of POTDR, at least aspect the PMD measurement.In order directly to measure the PMD at specific range z place, promptly accumulate PMD, the relation of second kind of POTDR use between those POTDR tracks (OTDR trace) that two or more tight alternate wavelength obtain, simultaneously do not need the birefringent characteristic of this optical fiber is carried out any supposition, also explicit or implicit expression integration that need be on length, the section of yet not missing does not have the problem of spin fiber yet, and is like that.Even if the PMD of circular birefringent fiber or polarization maintaining optical fibre (PMF) is also correctly measured.Different with the enforcement of first kind, for deriving PMD quantitatively, the enforcement of second kind does not need to introduce supposition and complex model.

Like this, accumulation PMD and the measurement of slope (speed that PMD changes with distance) accordingly thereof to changing with the distance z on the optical fiber that this second kind of POTDR can realize have promoted reliable recognition and quantitatively characterizing to those independent, short relatively " bad " sections mentioned above.

In second kind, the POTDR technology that is widely known by the people most depends on: have deterministic relation between OTDR track that obtains with a small amount of specific input SOP and output polarization analyzer axle, for example U.S. Patent No. 6,229, people's [14,15] such as 599 (Galtarossa) [16] and H.Sunnerud article is disclosed.This requires FUT is stable on the whole time period internal space of measuring all tracks.Unfortunately, in fact such stability can not be guaranteed, especially measures under the situation of mounted optical fiber.

In addition, second kind of known technology requires to use short pulse; The meaning of " weak point " is the long and coupling length of any section bat that pulse length is shorter than FUT.For the PMD in the measuring optical fiber correctly, and be not limited to have the short long optical fiber of clapping, these technology must be used usually the OTDR light impulse length less than about 10 nanoseconds.Unfortunately, Shi Ji OTDR does not have between a useful dynamic area the pulse of weak point like this.On the other hand, if use long light pulse, then only can measure and have the long long optical fiber of clapping, this limits these technology generally and goes to measure short distance and/or measure for a long time, or measures and have the long optical fiber of clapping long (typically little PMD coefficient).Thereby, carry out successful relatively PMD measurement though perhaps might use known technology and meet above-mentioned requirements, at present, its range of application and performance still are not enough to be used for the independent instrument of viable commercial.

In addition, the signal to noise ratio (snr) problem that causes thus that the use of short pulse has increased the weight of to be superimposed upon the so-called coherent noise on the OTDR track---coherent noise will increase when using short pulse---.This be because, the power of back reflective light just in time is not the power sum that each unit (dz) from this optical fiber sends.For coherent source,, between different backscattering sources, exist and interfere such as used laser of narrowband light source in the POTDR application.Be superimposed upon this interference or coherent noise on the ideal trajectory (power sum), the two all is inversely proportional to pulse width (or duration) and laser linewidth.It also can reduce in the following way: by increasing laser linewidth of equal value, be intrinsic laser linewidth etc., maybe can be by using " dither " or on wavelength, track being averaged, but this has reduced maximum detection amount PMD value, thereby also may limit maximum detection amount length, because PMD increases along with the increase of length.Say that roughly condition is PMD live width＜1 (wherein live width is an optical frequency unit); Otherwise useful POTDR signal will " be washed away " by unpolarizing.

Therefore, it is desirable to, having a kind of pulse of using its length to grow greater than the FUT bat to measure accumulation PMD quantitatively (has between high dynamic area, keep gratifying spatial resolution simultaneously) technology, and simultaneously the stability of FUT is not strict with, also need not optical fiber behavior (for example strong mode coupling) is supposed.

In a word, need a kind of new being used to characterize the method for this polarization-dependency characteristic of light path, this method is to the fibre movement that occurs under the field condition and the robust of perturbation own and do not require polarization optics device expensive and trouble.Preferably, this basic methods should be established several different embodiments, these embodiments especially be fit to the DGD in the narrow DWDM channel, under a plurality of wavelength DGD, PMD and carry out single-ended and/or double-end measurement with the accumulation PMD that the distance on the optical fiber link changes.

Summary of the invention

The present invention seeks to eliminate or alleviate at least the shortcoming of above-mentioned prior art, or a kind of replacement scheme is provided at least.

According to a first aspect of the invention, a kind of be used to measure at least one of light path (FUT) and the method for polarization-correlation properties are provided, this method is used near-end or near the optical input device of near-end that is connected to this light path, near and the near-end or the far-end that are connected to this light path, or the light output arrangement near-end or the far-end; This optical input device comprises the light supply apparatus of the light that is used to be supplied to the small part polarization, and the device that is used to control the input polarized state of light (I-SOP) of described partial polarization at least and described light is sent into this FUT; The light output device comprises the device that is used for extracting from this FUT corresponding light, be used to analyze the analytical equipment of the light that is extracted, and pick-up unit, described pick-up unit is used to detect the analysis light (A-SOP) corresponding at least one transmission axle of this analyser device, with the coherent light power of the transmission under each wavelength in each group that at least two group wavelength are provided, wherein each organizes the low wavelength (λ in wavelength _L) and high wavelength (λ _U) closely alternate, and wherein following three conditions are not all to be satisfied simultaneously:

A. (input) source and pick-up unit are in the same end of FUT;

B. in analysis and pick-up unit, only use a detecting device;

C. the light from light source mainly comprises transient pulse, and the space stretching of this pulse surpasses FUT and claps long 10 times;

And wherein said group to comprise a wavelength right, described wavelength in each group is to all corresponding little optical frequency is poor, and at described wavelength to defining a mid point wavelength, and wherein I-SOP and A-SOP are for each the described wavelength substantially constant in each described group, and wherein, between each described group, at least one among mid point wavelength, I-SOP and the A-SOP is different, and this method may further comprise the steps:

I. calculate at least one poor corresponding to the power parameter that records of each wavelength of described wavelength centering for each batch total in described at least two groups, described power parameter that records and described power analyzed and detected light subsequently are proportional, define the set of at least two power parameter differences that record thus;

Ii. calculate the mean square value of the set of described difference; And

Iii. calculate at least one polarization-relevant FUT characteristic as at least one predefined function of described mean square value, described predefined function depends on that the little optical frequency between the described wavelength is poor, and described wavelength is corresponding to each the right wavelength in described at least two pairs of tight alternate wavelength; And

Iv. export the value of described at least one polarization-relevant FUT characteristic.

For double-end measurement, described light output arrangement can be connected near the far-end or far-end of this FUT.

Preferably, DGD for the certain wave strong point measures, for example, and for narrow DWDM channel measurement, each described group wavelength that all comprises the mid point wavelength that has defined substantially is right, and described at least one polarization-relevant FUT characteristic is the Differential Group Delay (DGD) at described mid point wavelength place.

Described measured power parameter can be the standardization power T that calculated (v), described predefined function for little optical frequency poor (δ v) can express according to following differential formulas:

DGD (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{{< ΔT {(v)}^{2} >}_{SOP}}

Constant wherein

α_{ds} = \sqrt{\frac{9}{2}},

V is the optical frequency corresponding to described mid point wavelength.

According to a second aspect of the invention, provide the surveying instrument of at least one the polarization-correlation properties that is used to measure light path (FUT), comprised:

Optical input device, it is used to be connected near the near-end or near-end of this light path, and

Light output arrangement, it is used to be connected to the near-end or the far-end of this light path, or near-end or far-end near, extracting, to analyze and detect the light of at least a portion of having passed this FUT, and provide corresponding electric signal, and

Treating apparatus, it is used to handle the electric signal from the light output device, to determine described at least one polarization-correlation properties;

This optical input device comprises:

Light supply apparatus, it is used for supplying under each ripple of at least two set of wavelengths at least

The light of partial polarization, and

The SOP control device, it is used to control the input polarized state of light (I-SOP) of described partial polarization at least, and described light is sent into this FUT, wherein low described wavelength (λ L) in each described set of wavelengths and high described wavelength (λ _U) closely alternate, described group to comprise a wavelength right, described wavelength in each group is poor to corresponding little optical frequency, and at described wavelength to limiting a mid point wavelength, and the SOP of the light of being sent into and A-SOP be for each the described wavelength substantially constant in each described group, and wherein, between each described group, among mid point wavelength, I-SOP and the A-SOP at least one is different, and

The light output device comprises:

Extract and analytical equipment, it is used for extracting corresponding light from this FUT, and analyzes the light that is extracted, and

Pick-up unit, it is used to detect the analysis light (A-SOP) corresponding at least one transmission axle of this analyser device, with the transmission coherent light power of analysis light under each wavelength in each group that described at least two group wavelength are provided, the wherein low described wavelength (λ in each described set of wavelengths _L) and high described wavelength (λ _U) closely alternate, and wherein following three conditions are not satisfied simultaneously:

D. input source and pick-up unit are in the same end of FUT;

E. in analysis and pick-up unit, only use a detecting device;

F. the light from light source mainly comprises transient pulse, and the space of this pulse extends beyond FUT and claps long 10 times;

This treating apparatus is configured and can operates to be used for:

V. calculate at least one poor corresponding to the power parameter that records of each wavelength of described wavelength centering for each batch total in described at least two groups, described power parameter that records and described power analyzed and detected light subsequently are proportional, define the set of at least two power parameter differences that record thus;

Vi. calculate the mean square value of the set of described difference; And

Vii. calculate at least one polarization-relevant FUT characteristic as at least one predefined function of described mean square value, described predefined function depends on that the little optical frequency between the described wavelength is poor, and described wavelength is corresponding to each the right wavelength in described at least two pairs of tight alternate wavelength; And

Viii. export the value of described at least one polarization-relevant FUT characteristic, so that show, transmit or further handle.

According to a third aspect of the invention we, provide to be used for continuously and repeatedly producing two or more closely light source facilities of the coherent light of alternate wavelength, this facility comprises:

Optical gain medium;

At least two laser cavitys that comprise described optical gain medium, the part of their each laser cavity is shared in each chamber;

At least one output coupler, it allows to extract the part corresponding to light in the chamber of each described at least two laser cavity;

Beam splitter, it is used for this light is divided into separated portions at least two spaces, and each described at least two laser cavity is corresponding in described at least two parts at least one;

Multichannel tunable wave length bandpass filter device, it comprises at least two channels corresponding to different tight alternate wavelength, it can be operated with will be corresponding to each the light-receiving in the separated portions on described two spaces in channel separately at least, and can operate with the method for synchronization described channel is carried out wavelength tuning; And

The multi-channel optical retention device, it can be operated so that be no more than the described smooth partial continuous on it of inciding of separating on the space, and stops every other light part, and the parameter of described multi-channel optical retention device is depended in the selection of the light that is not blocked part.

Specify effectively for the U.S., according to an aspect of the present invention, a kind of method that is used to measure at least one polarization-correlation properties of light path (FUT) is provided, this method is used near-end or near the optical input device of near-end that is connected to this light path, near and the near-end or the far-end that are connected to this light path, or the light output arrangement near-end or the far-end; This optical input device comprises the light supply apparatus of the light that is used to be supplied to the small part polarization, and is used to control the polarized state of light (I-SOP) of described partial polarization at least and described light is sent into the device of this FUT; The light output device comprises the device that is used for extracting from this FUT corresponding light, be used to analyze the analytical equipment of the light that is extracted, and pick-up unit, described pick-up unit is used to detect the analysis light (A-SOP) corresponding at least one transmission axle of this analyser device, with the coherent light power of the transmission under each wavelength in each group that at least two group wavelength are provided, wherein each organizes the low wavelength (λ in wavelength _L) and high wavelength (λ _U) closely alternate;

And wherein said group to comprise a wavelength right, described wavelength in each group is poor to corresponding little optical frequency, and at described wavelength to limiting a mid point wavelength, and wherein I-SOP and A-SOP are for each the described wavelength substantially constant in each described group, and wherein, between each described group, it is different having one among mid point wavelength, I-SOP and the A-SOP at least, and this method may further comprise the steps:

Ix. calculate at least one poor corresponding to the power parameter that records of each wavelength of described wavelength centering for each batch total in described at least two groups, described power parameter that records and described power analyzed and detected light subsequently are proportional, define the set of at least two power parameter differences that record thus;

X. calculate the mean square value of the set of described difference; And

Xi. calculate at least one polarization-relevant FUT characteristic as at least one predefined function of described mean square value, described predefined function depends on that the little optical frequency between the described wavelength is poor, and described wavelength is corresponding to each the right wavelength in described at least two pairs of tight alternate wavelength; And

Also the U.S. is specified effectively, according to a further aspect in the invention, provides the surveying instrument of at least one the polarization-correlation properties that is used to measure light path (FUT), comprise:

This optical input device comprises

Light supply apparatus, it is used for being provided at the light of the partial polarization at least under each wavelength of at least two set of wavelengths, and

The SOP control device, it is used to control the polarized state of light (I-SOP) of described partial polarization at least, and described light is sent into this FUT,

The low described wavelength (λ in each described set of wavelengths wherein _L) and high described wavelength (λ _U) closely alternate,

Described group to comprise a wavelength right, and the described wavelength in each group is poor to corresponding little optical frequency, and at described wavelength to limiting a mid point wavelength, and

The SOP of the light of being sent into and A-SOP are for each the described wavelength substantially constant in each described group, and wherein, between each described group, it is different having one among mid point wavelength, I-SOP and the A-SOP at least, and

The light output device comprises:

G. pick-up unit, it is used to detect the described analysis light (A-SOP) corresponding at least one transmission axle of this analyser device, with the transmission coherent light power of analysis light under each wavelength in each group that described at least two group wavelength are provided, the wherein low described wavelength (λ in each described set of wavelengths _L) and high described wavelength (λ _U) closely alternate;

This treating apparatus is configured and can operates to be used for:

Xii. calculate at least one poor corresponding to the power parameter that records of each wavelength of described wavelength centering for each batch total in described at least two groups, described power parameter that records and described power analyzed and detected light subsequently are proportional, define the set of at least two power parameter differences that record thus;

Xiii. calculate the mean square value of the set of described difference; And

Xiv. calculate at least one polarization-relevant FUT characteristic as at least one predefined function of described mean square value, described predefined function depends on that the little optical frequency between the described wavelength is poor, and described wavelength is corresponding to each the right wavelength in described at least two pairs of tight alternate wavelength; And

Xv. export described at least one polarization-relevant FUT characteristic, be used for showing, transmit or further handle.

The preferred embodiment of aforementioned five aspects of the present invention and type are set forth in appended dependent claims.

According to---it only is illustrated in the mode of embodiment---detailed description that combines with accompanying drawing hereinafter to the preferred embodiments of the invention, aforementioned and other purposes, feature, aspect and advantage of the present invention will be more apparent.

Description of drawings

Both-end PMD measures

Fig. 1 is the general schematic view of simplification of several sections that is connected to the surveying instrument at tested optical fiber (FUT) two ends, and this instrument is used for this FUT is carried out double-end measurement, to determine DGD and/or average DGD and/or rms DGD;

Figure 1B is and the similar rough schematic view of Fig. 1, but instrument shown in Figure 1B uses the detecting device of tunable laser source, an input-SOP controller (scrambler), an output-SOP controller (scrambler), a polarizer/analyzer and a Measurement and analysis light;

Fig. 1 C is the rough schematic view with instrument like the instrument class shown in Figure 1B, but instrument shown in Fig. 1 C uses a coupling mechanism, a polarizer and two detecting devices; One of them detecting device is used to measure through the analysis light behind the polarizer, and another detecting device is used to measure and the proportional light of total Output optical power from FUT.

Fig. 1 D is the rough schematic view with instrument like the instrument class shown in Figure 1B, but instrument shown in Fig. 1 D has two detecting devices that are connected to coupling mechanism, in order to measure the power of two repetitions, so that reduce the uncorrelated noise in measuring;

Fig. 1 E is the rough schematic view with instrument like the instrument class shown in Fig. 1 C, but instrument shown in Fig. 1 E has single detector and an optical switch, this optical switch is used for detecting device optionally is connected to the analysis light that is used to measure from polarizer, and measure from coupling mechanism, with the proportional light of total Output optical power from FUT;

Fig. 1 F is the rough schematic view with instrument like the instrument class shown in Fig. 1 E, but coupling mechanism of instrument shown in Fig. 1 F and polarizer are replaced by polarization beam apparatus (PBS), and optical switch optionally is connected to single detector many (two) individual output port of this PBS;

Fig. 1 G is the rough schematic view with instrument like the instrument class shown in Figure 1B, but instrument shown in Fig. 1 G uses a PBS and two detecting devices;

Fig. 1 H is the rough schematic view with instrument like the instrument class shown in Figure 1, but instrument shown in Fig. 1 H has a polarimeter, and it is used to analyze and detect the light from FUT;

Fig. 1 I be with like the instrument class shown in Figure 1B based on the rough schematic view of the both-end PMD measurement/testing tool of wideband light source, but instrument shown in Fig. 1 H uses a wideband light source that light is provided, and use a tunable optic filter (between polarizer and detecting device), so that only can detect the light in the little spectrum width that passband wavelength with this wave filter is the center;

Fig. 1 J be with like the instrument class shown in Fig. 1 I based on the rough schematic view of the both-end PMD measurement/testing tool of wideband light source, but instrument shown in Fig. 1 J uses a dispersion element (multi-channel filter) and a multi-channel detector array, and this multi-channel detector array is measured simultaneously or in a short time through the analysis light behind the polarizer.

Single-ended overall PMD measures

Fig. 2 is corresponding with Fig. 1, but Fig. 2 is the rough schematic view that is used for the measurement testing tool of single-ended overall PMD measurement;

Fig. 2 B to 2G is corresponding with Figure 1B to 1G respectively, shows corresponding single-ended surveying instrument, and wherein two parts of this surveying instrument are in the near-end of FUT together.

Single-ended accumulation PMD measures

Fig. 3 is the rough schematic view that has embodied the Polarization-Sensitive optical time domain reflectometer (POTDR) of one aspect of the present invention;

Fig. 3 A is the rough schematic view that has embodied the Polarization-Sensitive optical time domain reflectometer of one aspect of the present invention;

Fig. 3 B is the Polarization-Sensitive optical time domain reflectometer that has embodied one aspect of the present invention;

Fig. 3 C is the Polarization-Sensitive optical time domain reflectometer that has embodied one aspect of the present invention;

Fig. 4 A is a process flow diagram, and its figure has shown the light source and the operation of importing the SOP controller of the both-end PMD surveying instrument of Fig. 1 C and 1G;

Fig. 4 B is a process flow diagram, and its figure has shown the analyzer of both-end PMD surveying instrument of Fig. 1 C and 1G and the operation of detecting unit;

Fig. 4 C is a process flow diagram, and its figure has shown one group of power (data) obtaining step of the process flow diagram of Fig. 4 B;

Fig. 4 D is a process flow diagram, and its figure has shown a power (data) obtaining step of the process flow diagram of Fig. 4 C;

Fig. 5 A has demonstrated the several sections of the process flow diagram that the single-ended PMD of pictorial image 2C and 2G measures;

Fig. 5 B is one group of power (data) obtaining step of the process flow diagram process flow diagram that shown Fig. 5 A;

Fig. 5 C is a power (data) obtaining step of the process flow diagram process flow diagram that shown Fig. 5 B;

Fig. 6 A is a process flow diagram, and it has shown the operation of the POTDR of Fig. 3;

Fig. 6 B is a process flow diagram, and it has shown the obtaining step of a track of the process flow diagram of Fig. 6 A;

Fig. 7 is that figure has shown a tunable modulated optical light source synoptic diagram;

Fig. 7 A has shown an embodiment based on the synoptic diagram of the tunable modulated optical light source of SOA;

Fig. 8 A has shown a synoptic diagram that can be used to the tunable light-pulse generator that has time delay of single-ended overall PMD measurement and single-ended accumulation PMD measurement;

Fig. 8 B shown that another can be used to that single-ended overall PMD measures not with the synoptic diagram of the substituting tunable light-pulse generator of time delay;

Fig. 8 C has shown that another can be used to the tunable light-pulse generator that substitutes that single-ended overall PMD measures and single-ended accumulation PMD measures;

Fig. 9 A is the rough schematic view of a lasing light emitter, and this lasing light emitter has been adjusted to guarantees that the light that is sent has high-polarization (DOP);

Figure 10 A and 10B can be used to that single-ended overall PMD measures and the schematically showing of the substituting tunable light-pulse generator of single-ended accumulation PMD measurement.

Embodiment

In the accompanying drawings, identical or similar member has identical reference number in different figure, has the symbol of an indication difference under the suitable situation.

Various aspect of the present invention, and their enforcement separately are all based on identical basic theory.The embodiment of these aspects can be advantageously used in: at narrow optical channel or in the wide range of wavelengths of appointment, to PMD or depend on that the double-end measurement of the DGD of wavelength, single-ended overall PMD are measured, single-ended accumulation PMD measurement and other related variants.

In each preferred embodiment of the present invention of Miao Shuing, three major parts will be arranged usually hereinafter, i.e. (i) input optical controller, (ii) analyzer or detecting unit and (iii) simulation and digital processing element are together with one or more control modules.In so-called both-end case, the input optical controller will be positioned at the near-end of FUT, and analyzer and detecting unit and be also to comprise simulation and digital processing element easily will be positioned at the far-end of FUT.Be in first control module control input optical controller of FUT near-end, be in the second control module control analysis device and detecting unit and the simulation and the digital signal processing unit of FUT far-end.In the single-ended case of indication, all be under the situation of FUT near-end at all members of this surveying instrument, these two control modules can be combined into single control module.

All will have above-mentioned three parts or section though embody each instrument of the present invention, but measure type according to these three kinds different PMD, promptly the overall PMD of (i) both-end measures, (ii) single-ended overall PMD measures and (iii) single-ended accumulative total PMD measurement, aspect the configuration difference on many details will arranged.

Like this, the input optical controller will comprise a light source of polarization at least in part---for example tunable laser or wideband light source, and an input SOP controller---it is used for the SOP of this light of control before the light from this light source is admitted to FUT.Except an output SOP controller, analyzer and detecting unit can also comprise a polarizer and a detecting device, or a PBS and two detecting devices, or a coupling mechanism, a polarizer are like that together with two detecting devices.At the input light source is under the situation of broad band source, and analyzer and detecting unit also can comprise a tunable optic filter, is used to select optical frequency (alternatively, but have inferiorly slightly, input light source can comprise is tunable optic filter so).Simulation and digital processing element can comprise a digital acquiring unit, sample and ask an averaging unit and a data processor unit, in this sampling and ask and carry out analog-digital conversion in the averaging unit.

Use single-ended mensuration, by analyzing back reflective light from the strong local reflection (for example Fresnel reflection, Bragg reflector or the like) of FUT far-end, can obtain overall PMD value, therefore can advantageously use a long pulse, because all back reflective light all are derived from the local reflection in fact, rather than from the rayleigh backscattering that distributes along pulse length.Also can preferably use two tight alternate wavelength to measure.Yet, measure accumulation PMD in order to use single-ended mensuration, must analyze the OTDR track that changes along with fiber lengths, therefore can preferably use a short optical pulse, so that obtain POTDR track clearly, this POTDR track does not suffer to depolarize with respect to the SOP of " back along " the unsuitable space of causing that develops because of the pulse " forward position " due to the PMD.

In addition, typically, for once obtaining, the accumulation PMD " curve " that changes along with the fiber lengths that requires to measure has one approximate " continuous " to be increased.Because, for a given tight alternate wavelength gap, there are a maximum PMD value (owing to saturated) and the minimum PMD value (owing to detection sensitivity) can be measured, thereby also preferably, send into light pulse of (for example three or four) the closely alternate wavelength that has two or more.In this way, the PMD value of the measurement of being carried out under different tight alternate wavelength intervals can be in processing be arrived together by " stitchings ", thereby measurable minimum and maximum PMD value can be strengthened significantly.

Measure for both-end PMD, laser instrument must be provided with or modulate its optical frequency, to produce two or more closely alternate wavelength at different time.

Be used for now with reference to Fig. 1 to 3C that both-end PMD of the present invention measures, single-ended overall PMD measures and the preferred embodiment and the instrument configuration of three kinds of main PMD mensurations that single-ended accumulation PMD measures, with and revise, change and replacing, be described.

Both-end PMD measures

In the description hereinafter at both-end PMD measurement, term " modulating light pulse " refers to such propagates light: it distinguished aspect one or more following with at least some other pulses in the time interval that limits: characteristic wavelength, feature average power, characteristic pulse duration, feature stack amplitude or than the phase modulation (PM) under reciprocal high a lot of frequency in duration of pulse, its duration the feature extinction ratio, at feature duration or any other measurable distinctiveness attribute of the sampling of obtaining light described in the processing.

In first preferred embodiment of the present invention shown in Figure 1, test/measurement the facility that is used for both-end DGD/PMD measurement comprises: input optical controller device 42, it is positioned near the proximal end or near-end of FUT 18, and is connected to by connector 16A near the proximal end or near-end of FUT 18; And analyzer and pick-up unit 44, it is positioned near the far-end or far-end of FUT 18, and is connected to by connector 16B near the far-end or far-end of FUT 18.Input optical controller device 42 comprises light source 12 and input SOP control device 14A (be called as the I-SOP controller easily or disturb deflection device), and it controlled the SOP of this light before will sending into FUT18 via connector 16A from the light of light source 12.

Under the not high situation of the degree of polarization (DOP) of light source 12,, can improve DOP by polarizer 19 (for example polarizer, polarization beam apparatus or the like) is inserted in the light path in light source 12 downstreams.Yet, if do not use polarization maintaining optical fibre (PMF) between light source 12 and the polarizer 19, has the additional polarization adjuster 13 (the normally Polarization Controller that " in producer, is provided ") of necessary interpolation so, shown in Fig. 9 A, so that make transmission through the approximate maximization of the power of polarizer 19.It should be noted that polarizer 19 can be identical with the polarizer of the particular that is used for one-sided measurement (20,20A, 20C), as illustrated in Fig. 2 B-G and 3A and 3B.

The wavelength of first (input) control module 30A control tunable laser source 12A, and the setting of input I-SOP controller 14A are used for specifically, and the SOP to this light before the light from light source 12 is admitted to FUT 18 disturbs inclined to one side control.

Analyzer and pick-up unit 44 comprise output SOP controller (A-SOP) 14B (be called as the A-SOP controller easily or disturb deflection device), are polarization Discr.s 20 thereupon, and pick-up unit 22.---for example optical attenuator---can be inserted into if pick-up unit 22 can not correctly be measured high luminous power, output control device (not shown) so, before the light that extracts from FUT 18 is applied to pick-up unit 22 this light is decayed.The purpose of this optical attenuator is to guarantee that the light level of far-end can be not high to making pick-up unit 22 " saturated " potentially or presenting nonlinear effect.If, for example on the short fiber link, carry out to measure---the overall decay of wherein being introduced by optical fiber is little, just may be such situation.For the long-chain road, optical attenuator will be set to minimal attenuation usually.

Simulation and digital processing element 40 comprise sampling and ask averaging unit 32 and data processor means 34, have the display equipment 36 that is used for display result alternatively.Member of analyzer and pick-up unit 44 (except the polarization Discr.) and simulation and digital processing element 40 are controlled by the second output control unit 30B.

Working in coordination with down of control module 30B, sampling and/or averaging circuit 32 use internal simulation-digital quantizer that the electric signal of the correspondence that changes along with the time of coming self-

detector

22B and 22C is sampled in a known way, the signal of being sampled is through time averaging, so that corresponding digital level to be provided in its part duration.This part is selected, to avoid transient effect and/or the bandwidth constraints effect in the power that is detected, polarization and/or wavelength, described these effects are caused by light supply apparatus 12, I-SOP controller 14A, the analytical equipment that comprises A-SOP control device 14B and polarization Discr. device 20 and/or any (pulse) signal distortion that is caused by the analog electronics bandwidth constraints.

Averaging of income power is used for drawing FUT18 in DGD value under the specific wavelength or the PMD value in the provision wavelengths interval by data processor 34, as will be hereinafter according to a particular aspect of the invention in more detail as described in.

The various differences of the both-end instrument of Fig. 1 are configured in shown in Figure 1B to 1J, will describe briefly now.The something in common of the instrument configuration that Fig. 1 to 1H describes is that they all use tunable laser source, and the instrument configuration that Fig. 1 I and 1J describe is used wideband light source and tunable optic filter.

Like this, in each shown in Fig. 1 to 1H " both-end " instrument, light source 12A comprises tunable optical modulated laser source 12A, its output, depend on the circumstances, be coupled to polarization maintaining optical fibre (PMF) or single-mode fiber (SMF), so that modulating light pulse is sent into tested optical fiber (FUT) 18 via (input) polarization state (I-SOP) control device 14A and input connector 16A.The output light that extracts from FUT 18 is polarized Discr. 20 and analyzes, and described analyzed light is measured in a period of time---in this time period, from two different wave lengths, i.e. λ of light supply apparatus 12 _L ^(k)And λ _U ^(k)Under light detected in succession, described two different wave lengths are closely alternate each other.

The key distinction between these different configurations is analyzer and pick-up unit 44.To this, in the analyzer and pick-up unit 44 of the instrument shown in Figure 1B, the polarization Discr. comprises linear polarization 20A, and pick-up unit comprises single detector 22A.

Fig. 1 C shows and instrument like the instrument class shown in Figure 1B, but difference is that instrument shown in Fig. 1 C has two detecting

device

22B and 22C and a coupling mechanism 21 that is inserted between A-SOP controller 14B and polarization Discr. (polarizer) 20A.Detecting device 22B is connected to polarizer 20A, and measures the described analyzed light from polarizer 20A; Detecting device 22C is directly connected to coupling mechanism 21, and the proportional light of general power of the light of measuring and being extracted from FUT 18.The SOP of the light that is extracted is by A-SOP controller or scrambler 14B conversion, and this light is coupled device 21 separated into two parts thereupon.The polarizer that is connected to one of two output terminals of coupling mechanism 21 is connected to the first detecting device 22B and measures described analyzed light from polarizer; Be connected to the second detecting device 22C of another output terminal of coupling mechanism 21, measure and the proportional power of total Output optical power from FUT.This light can be similar to simultaneously and be detected by detecting device 22B and 22C.Yet, it should be noted that always not to be necessary the real light of side by side being analyzed with two detecting

device

22B and 22C detection, but the light that can analyze in different slightly time detecting.

Instrument class shown in instrument shown in Fig. 1 D and Fig. 1 C seemingly, difference is polarizer 20A and coupling mechanism 21 transpositions, two detecting

device

22B and 22C are connected respectively to each (two) individual output terminal of output coupler 21, to measure the power of two repetitions.

Instrument class shown in instrument shown in Fig. 1 E and Fig. 1 C seemingly, they all comprise coupling mechanism 21 and polarizer 20A, but the difference be that instrument shown in Fig. 1 E only has a detecting device 22A.The optical switch 23 of Be Controlled unit 30B control optionally is connected to the output terminal of coupling mechanism 21 and the output terminal of polarizer 20A with the input end of detecting device 22A, with measure respectively the light analyzed and with the proportional luminous power of total Output optical power from FUT 18.

Instrument class shown in instrument shown in Fig. 1 F and Fig. 1 E is seemingly distinguished and is that instrument shown in Fig. 1 F uses single detector 22A and an optical switch 23, but have PBS 20C rather than linear polarization.Control module 30B makes switch 23 that detecting device 22A optionally is connected to each (two) individual output port of PBS 20C, to measure the light of being analyzed from each port.

Because optical switch 23 is used to importing same detecting device from coupling mechanism 21 and polarizer 20A (Fig. 1 E) or from the output light of two light paths of PBS 20C (Fig. 1 F), so can be detected in the different time from the light of these two different light paths.This will allow to use an only detecting device (and related electron device), keep many and two advantages that detecting device is related of use simultaneously.Certainly, the cost that being reduced in of the cost related with only using detecting device will be introduced into optical switch to a great extent to be increased is offset, and has the cost on the Measuring Time.

Instrument class shown in instrument shown in Fig. 1 G and Fig. 1 F seemingly, but difference is that instrument has omitted switch shown in Fig. 1 G, and two detecting device 22B and 22C be connected respectively to each (two) individual output port of PBS 20C, to measure the light of being analyzed from this each port.By A-SOP controller or scrambler 14B conversion, this light is decomposed into two components by PBS 20C thereupon from the SOP of the light of the far-end of FUT 18, and the SOP that it has quadrature typically is the linear SOP that 0 degree and 90 is spent relative orientations.The first detecting device 22B is connected to one of two output terminals of PBS 20C, to receive one of these quadrature components; Another output terminal of PBS 20C (with respect to the light from FUT 18) is connected to the second detecting device B 22C, to receive another quadrature component.In case through suitably calibration, so that relative detector efficiency, wavelength dependency etc. are taken into account, as hereinafter describing, come the summation of the power that is detected of self-detector 22B and 22C (usually to be called as the Stokes parameter S respectively with total incident (promptly not analyzing) power ₀) proportional.This light can be by detecting device 22B and the approximate detection side by side of 22C.

It will be appreciated that at polarization Discr. 20 to comprise under the situation of polarizer 21A and coupling mechanism 21 (Fig. 1 C) that the detecting device 22C that is connected to coupling mechanism 21 receives unpolarized relevant light.

Instrument class shown in instrument and Figure 1B shown in Fig. 1 H seemingly, but difference is that the analyzer of instrument shown in Fig. 1 H and pick-up unit 44 comprise polarimeter 45, the input end of polarimeter 45 is connected to FUT 18 via connector 16B, and output terminal is connected to sampling and asks averaging unit 32.Polarimeter 45 Be Controlled unit 30B control is to analyze and to detect the light that receives from FUT 18.

The preferred embodiment of use wideband light source 12B of the present invention rather than tunable laser source 12A is described now with reference to Fig. 1 I and 1J.Shown in Fig. 1 I shown in measurement/test facilities and Figure 1B and similar with reference to the described measurement facility of Figure 1B, but difference is the input optical controller device 42 of measuring facility shown in Fig. 1 I and comprises polarized broadband light sources 12B rather than tunable laser source, and analyzer shown in Fig. 1 I and pick-up unit 44 are different with analyzer shown in Figure 1B and pick-up unit 44, because the former has the tunable optic filter 27 that is inserted between polarizer 20A and the detecting device 22A.Tunable optic filter 27 Be Controlled unit 30B control.

Should recognize, tunable optic filter 27 can optionally be placed in the output terminal of FUT 16B and any position in the light path between the detecting device 22A, keep next-door neighbour's control module 30B simultaneously, and be not limited to shown in Fig. 1 I, be placed between polarizer 20A and the detecting device 22B.In fact, more generally be that tunable optic filter 27 can be placed in any position between broad band source 12B and the detecting device 22A.Yet, place the input optical controller 42 of FUT 18 proximal ends can cause controlling and synchronous difficulty in this wave filter, because the communication between the control module 30B of the tunable optic filter 27 of FUT proximal end and far-end is with difficulty.

In the embodiment of Fig. 1 I, if the intrinsic DOP of this wideband light source is not too high, so, by before will sending into FUT 18, making this light through polarizer from the light of broadband light source 12B, and regulate the SOP of this incident light, can obtain " polarization well " broadband light (seeing Fig. 9 A).In this case, an additional polarization adjuster (being Polarization Controller) and a polarizer (seeing Figure 10 A, 10B and 2D) will be inserted between wideband light source 12B and the I-SOP controller 14A.This Polarization Controller will be regulated the input SOP of light, to obtain approximate maximum Output optical power from this polarizer.

Instrument class seemingly shown in instrument shown in Fig. 1 J and Fig. 1 I, but difference is that tunable optic filter 27 is replaced by spectrometer device or multichannel filtering apparatus, dispersion element 27A for example based on the wavelength separator of grating, is used for separating these different optical wavelength with the variation of angle specifically.Single detector is replaced by and is used for the approximate pick-up unit that side by side detects the power of the light under these wavelength, for example multi-channel detector array 22D or similar device.Alternatively, detector array can be replaced by several optical fiber pigtail formula photodetectors---and it can be connected to fiber array and be in the light of different spatial with detection, or the light with different optical wavelength that will be in different spatial is simply sent into different light sensation detectors.Though this design has higher cost, it can measure DGD or PMD apace.

Preferably, in " both-end " surveying instrument shown in Fig. 1 to 1J, between the control module 40A of the control module 40B of FUT18 far-end and proximal end, there is not " up " communication.Control module 30B comprises software or firmware, these softwares or firmware allow control module 30B to determine that from following information the specific detected modulating light pulse that extracts is whether corresponding to the highest, minimum or in due course from FUT 18, it is medium tight alternate wavelength, described information is input optical controller 42---aptly, under the control of control module 30A---be encoded to the information on the optical signalling.

Above-described preferred embodiment aspect those principles of the present invention is being common usefulness.Yet, corresponding to each the details of preferred embodiment of these principle aspects, comprising their details of operation, will in next sub-chapters and sections, be described in more detail.

In the following description, term " modulating light pulse " refers to such propagates light: it distinguished aspect one or more following with at least some other pulses in the time interval that limits: characteristic wavelength, feature average power, characteristic pulse duration, feature stack amplitude or than the phase modulation (PM) under reciprocal high a lot of frequency in duration of pulse, its duration the feature extinction ratio, at feature duration or any other measurable distinctiveness attribute of the sampling of obtaining light described in the processing.In the linguistic context of more detailed description, the implication of " modulating light pulse " will be more clear hereinafter.

The DGD of certain wave strong point measures

In a narrow DWDM channel, above wavelength (λ in the measured channel _Mid) DGD that locates is often also unrealistic, because the optical frequency of the wavelength that these are closely alternate may account for available optics passband most at interval, therefore the measurement at another mid point wavelength place may cause the excessive decay of these two tight alternate wavelength experience, polarization-dependent loss and other to cause measuring unreliable or infeasible ill-effect.(as will be hereinafter in more detail as described in, very little optical frequency may be not enough to allow the measurement of little DGD value at interval.Yet), substantially, when the PMD of FUT relatively little---for example less than 0.2-0.5ps---time, DGD in little band in the range of wavelengths (such as 30GHz) can show little variation, although thereby usually still the expectation DGD that obtains each wavelength place obtain average DGD or rms DGD in this little channel wavelength interval.

The DGD measurement that also it should be noted that the certain wave strong point is not limited to " in the band " and uses, such as the optical link in the whole DWDM channel is tested.

Note, measure for the DGD in " special use " DWDM channel, the i.e. measurement of always carrying out in approximately uniform certain wave strong point, light supply apparatus 12 be not must the significantly tunable or non-constant width of bandwidth light source, only requiring it can launch with aforementioned " specific wavelength " is coherent light under two different tight alternate wavelength at center.Yet, measure to use for majority, expectation light supply apparatus 12 is tunable or the light source bandwidth is very wide, so that in several other DWDM channel wavelengths any one measured---for example in communication C and/or L band.The more detailed description of the operation of the preferred embodiment of this tunable optical source or wideband light source device will provide in sub-chapters and sections after a while.

As above describing in " background technology " chapters and sections, DGD can be along with time and/or changes in environmental conditions and is changed.Use for many measurements, the speed of this measurement (" renewal rate ") is not crucial.Therefore, for cost consideration, advantageously, be input-SOP controller 14A and the cheap polarization scrambler of analytical equipment selection.Being suitable for the two a example of low cost S OP scrambler of I-SOP and

A-SOP controller

14A and 14B is, at the U.S. Provisional Patent Application No.60/996 that owns together, on November 26th, 2007 submitted to, is described in 578.

The actual SOP of the light that comes out from input I-SOP controller 14A generally is unknown, but experience " continuous sweep ", i.e. change slightly between several groups of closely alternate wavelength, so that in the sufficiently long period, usually corresponding to in the short time interval of obtaining that reliable DGD measures, this SOP will be similar to and cover Poincare sphere equably.

The output A-SOP controller 14B that is positioned at the far-end of FUT 18 also changes the SOP of the light that comes out from FUT 18 in the mode that is similar to input I-SOP controller 14A lentamente, although change speed on the whole separately and inequality, and from I-SOP controller 14A and A-SOP controller 14B come out only incoherent.

More specifically, for specific measurement sequence k, control module 30B makes light signal, and---its polarization Discr. of being got involved (such as polarization beam apparatus (PBS) or polarizer) is analyzed---measured in portion of time---in this portion of time, two different wave lengths, i.e. λ from light supply apparatus 12/12A _L ^(k)And λ _U ^(k)Under light detected in succession, described two different wave lengths are closely alternate each other, in described portion of time, the SOP approximately constant that comes out from I-SOP controller 14A and A-SOP controller 14B respectively, and form k SOP to (I-SOP ^(k), O-SOP ^(k)) (preferably, aforementioned part is less than 50% of " physics " pulse length, and reason will be explained hereinafter.)。This is restricted to the mean value of the actual wavelength of modulating light pulse to the mid point wavelength of modulating light pulse, promptly

λ_{mid}^{(k)} = (λ_{L}^{(k)} + λ_{U}^{(k)}) / 2 .

(understand with being convenient to for convenience, mark L and U refer to respect to the mid point wavelength X _Mid ^(k)Wavelength " low " and " height " value.)

Measured analysis light signal is being applied to data processor 34 with before carrying out processing subsequently, is sampled and asks equilibration device 32 to be converted to electric signal and also then be digitized.

From a tight alternate wavelength Conversion to other process, temporarily extinguished for example about 40 μ s from the light of light supply apparatus 12A, be significantly shorter than the typical reaction period of the DWDM channel equalizer in many optic networks during this period of time.The pulse that the Be Controlled unit 30B of accurate period that this extinguishes is used for discerning subsequently is corresponding to high wavelength or corresponding to low wavelength.

Above-mentioned measurement sequence is repeated at K different group, and each is organized corresponding to a different slightly I-SOP and A-SOP.In practice, for continuous SOP scan mode, K should be greater than 1000, ideally greater than 10,000, to obtain gratifying result.

The photoemissive time period corresponding to each closely alternate wavelength place is not crucial especially, but clearly, the long duration will cause the long overall measurement time of the method.Have been found that the good compromise between Measuring Time and the restriction of optical source wavelength switch speed is the period of about 1ms.

If can not roughly know the DGD value that will measure, thereby so closely the alternate right optical frequency difference of wavelength just might not allow accurately to measure high DGD value too greatly, thus perhaps too little do not allow accurately to measure hang down the DGD value.Under such situation, ideal situation is only to use a limited number of K value to carry out preliminary rough DGD to estimate.(it should be noted that with continuous SOP mode, for rough measure, K must be still big relatively, for example greater than 500, yet if use substituting " macroscopical stride (macroscopic-step) SOP selection " mode, as mentioned below, K can be a much smaller value, for example is approximately equal to 10.) continue after, depend on the result, the interval of closely alternate wavelength can be conditioned, and keeps the mid point wavelength to be in identical value simultaneously.Yet as mentioned above, in---it can for example only have the available passband width that is approximately equal to 35GHz---, the increase wavelength interval is always unfeasible at narrow DWDM channel.

Another approach that is used for the optical frequency difference between " adaptive " closely alternate wavelength is, uses plural closely alternate wavelength in each group, wavelength between be not wait at interval.As mentioned above, if preliminary DGD estimation indication wavelength interval should be different, so only need and to arrive corresponding to the initial closely alternate right mid point wavelength of wavelength corresponding to the right mid point wavelength shift of " optimum " closely alternate wavelength.Such approach is suitable for optimized light source device 12, and its embodiment will be described in more detail below.

Advantageously, in order to estimate and the The noise in the compensating measure partly that at each group, carry out " duplicate measurements " at two identical tight alternate wavelength places, when not having noise, these duplicate measurementss are equal to " original " in principle substantially and measure.In practice, such noise may be derived from any combination of the following: ASE noise (from the optical amplifier of getting involved in the optical fiber link), polarization noise, light source power fluctuation or the like.Utilizing this technology to improve measurement sensitivity will be described in detail hereinafter.

Yet, should note, be easily, in this preferred embodiment, actual transmissions distinct " physics " repetition pulse not, but in acquisition process by in being different from the time portion of carrying out that part of time that " initially " measure, " physics pulse " (corresponding to period of Laser emission specific wavelength) being sampled, carry out the function of equivalence.Therefore, in a preferred embodiment, each " physics pulse " comprises two " optical modulation pulses "

Will be described in more detail below following computing method: by these computing method, the data of obtaining like this can be converted into reliable DGD and measure, and comprise under the situation that has significant ASE noise.

Use RMS or all DGD measurements of DGD (λ) measurement that repeats

By in the range of wavelengths of regulation, repeatedly using the method for the DGD of above-mentioned measurement certain wave strong point of the present invention, can estimate the polarization mode dispersion (PMD) (according to the arbitrary of " rms " or " on average " PMD definition or the two) of optical fiber link by DGD with wavelength variations.Preferably, these wavelength should be similar in the range of wavelengths that is evenly distributed in regulation.

Consider the overall measurement time, advantageously, the continuous SOP scanning that to describe in above " summary of the invention " replaces with " macroscopical stride SOP selects ", promptly wherein I-SOP controller 14A is provided with different input and output SOP with A-SOP controller 14B with pseudo-random fashion, so that those points that are used to SOP is indicated on the Poincare sphere routinely are evenly distributed on the surface of described ball, and though this distribution be at random or some lattice uniformly.An example of the suitable commercially available controller that is used for using like this is General Photonics Model PolaMight ^TM(multi-functional Polarization Controller).

As mentioned in above measuring about the DGD of certain wave strong point, situation usually is, closely the alternate right optical frequency difference of wavelength is, for example, thereby does not allow to measure exactly high DGD value too greatly, thereby or too low do not allow to measure hang down the DGD value.Under such situation, may it is desirable to, use the method, but carry out preliminary rough DGD estimation with limited K value (for example 10), then, depend on the result, change the interval of this tight alternate wavelength.Note, in this case---at rms or average DGD is in the wavelength coverage of regulation under calculated situation, and the measurement that has the different optical difference on the frequency hereto often needn't keep strict identical mid point wavelength.The final DGD on wavelength on average can consider the wavelength that these are different slightly.

A method for optimizing implementing the method with the preferred embodiment of light supply apparatus 12 will be described now.(for for simplicity previously described, " repetition pulse " method described in the DGD of certain wave strong point measured above our supposition did not have to utilize." middle wavelength " described herein method can be summarised as easily comprises " repetition pulse " method.) at first, input light control device 42 has the group of two light pulses for each, will have wavelength (λ ₁₁)---it is in the height and the low wavelength (λ of this group _1U, λ _1L) in the middle of and do not wait at interval with the two---the 3rd additional optical pulse send into FUT 18.Input SOP 14A and output SOP 14B be these three light pulses difference approximately constants for all.The detected system and device of all these three pulses of being analyzed 22 detects, and is identified by their " extinguishing the phase " separately, as mentioned described in the DGD of the certain wave strong point measurement.These three aforementioned lights pulses are corresponding to three different optical frequency subtractive combinations (in contrast thereto, two different tight alternate wavelength are certainly corresponding to only a possible optical frequency is poor), thereby the overall measurement time has only increased about 50%.Use hereinafter computing method in greater detail, can be in the range of wavelengths of regulation, the DGD that carry out noise and/or sensitivity optimization at different approximate alternate equably (mid point) wavelength places measures.

Should note, if use the remarkable asymmetric distribution of identical DGD (λ) quantity, so, by the conspicuous mode of those of ordinary skills is directly revised this method, the PMD value still can be calculated, but this PMD value generally will be reliable not as the PMD value that obtains with the approximate wavelength that distributes equably.

For fear of between input optical controller device 42 and analyzer and pick-up unit 44, using complicated communication, it is desirable to, (for example C is with for the range of wavelengths of stipulating, from 1530 to 1565nm), the selection of the tight alternate mid point wavelength that wavelength limited that tunable laser source 12A (Fig. 1 (B-H)) or tunable optic filter 27 (Fig. 1 I) are generated can be scheduled to.In this way, do not need to communicate by letter clearly and inform the digital value of the wavelength of sending into, because it is these values can be by drawing the simple coded message of control module 30B in the fall time, as indicated above.Yet, it is desirable to, initial " ready " signal is sent from input optical controller device 42, to begin to measure sequence.Equally, this signal can be encoded in the light of sending into FUT via the phase of extinguishing, or by other simple pulse frequency modulation.

In case one group of DGD (λ) value is obtained as described above, the different DGD value that so just is easy to use the canonical statistics definition to be obtained in the range of wavelengths of regulation is calculated the arbitrary of rms DGD and average DGD or the two.Notice that such measurement is particularly useful, directly do not measure PMD because most current commercial sources does not allow to use rms and on average defines the two.

RMS DGD measures (not being with individual DGD (λ) to measure)

Following metering system can be applied to the direct measurement to the rms DGD in the provision wavelengths interval (i.e. the PMD that defines according to rms).If do not need MeasureDGD is with the information of wavelength variations, and the RMS mensuration that the DGD (λ) that has realized in this respect repeating than aforementioned use so of the present invention measures is wanted faster PMD measuring method (for identical overall accurately level).In addition, do not need the actual value of the wavelength that " knowing " be transmitted (whether only to need to know this wavelength owing to analyze and detect optical controller device 44 corresponding to " height ", " low " or one or more " centre " wavelength), so do not need to use predetermined wavelength maybe definite " beginning " signal of this measurement, simplified measurement process thus.

Become reliable DGD to measure the computing method of (being included under the situation that has significant ASE noise) in order to the data-switching that will obtain like this, except " centre wavelength " in being similar to the interval that is evenly distributed in defined goes up (definition of centre wavelength is seen you then), on different input SOP and output SOP, and outside asking on average to the individual bulk measurement that each tight alternate set of wavelengths is done, measure roughly the same with above-mentioned DGD in the certain wave strong point.Ideally, but not necessarily, the selection of mid point wavelength be accurate at random, or do not arrange at least by the wavelength increasing or decreasing.The computational details will be described below.

Rms that the DGD (λ) that repeats for above-mentioned use measures or average DGD measure, and advantageously, send into plural different tight alternate wavelength in each set of wavelengths, so that can use optimum optical frequency at interval in computing.

Before the measuring process more than describing in more detail aspect these,,, should notice that such theory is not restrictive with the Theory of Interpretation basis in order to promote and to understand such operation.

Use the RMS DGD of quick length scanning to measure

An alternative measuring the rms DGD in the range of wavelengths of stipulating is, use scanning tunable laser (or combination of polarization broad band source/tunable narrow bandpass filter) fast, wherein I-SOP and A-SOP arbitrary or the two remaining unchanged or do not change fully almost in scanning process.If detection electronics is enough quick, this " spectrum obtaining step " will provide the transmission coherent light power data of the polarization-analysis of the quasi-continuous detection that changes along with optical frequency so.In data analysis subsequently, can select any required tight alternate wavelength stride, and to select from the average DGD value of different wavelength to determining with similar mode mentioned above.Certainly, if I-SOP and A-SOP change in scanning process, this will further improve the degree of accuracy of this measurement so, as long as in this scanning, between any two tight alternate wavelength, I-SOP and A-SOP keep not marked change.In addition, repeat this process with a plurality of scannings and can further improve degree of accuracy certainly.

Various modifications to both-end PMD measurement mechanism

The present invention has included the various modifications of the both-end PMD shown in Fig. 1-1H being measured embodiment.For example, if from the light of light supply apparatus 12 is not polarization well, be that different wavelength has different SOP, make this light through a polarization adjuster (being Polarization Controller) 13 (seeing Fig. 9 A) so, to produce maximum output light from polarizer by the incident SOP that regulates light, thereby make maximum luminous power through I-SOP controller 14A, described polarization adjuster is connected respectively to tunable pulsed laser source 12 and polarizer 19 by non-polarization maintaining optical fibre.

Though these modifications can be used discretely, particular of the present invention can comprise several such modifications.

Those of ordinary skills need not the relative sensitivity that excessive experimental skill just can be used to this process calibrate two detecting

device

22B and 22C, the loss that coupling mechanism that comprises intervention by mentioned earlier or the like causes wherein can be measured with reference to the used both-end PMD based on polarized light source of the embodiment of figure 1G.That is to say, it will be appreciated that in the embodiment of Fig. 1 C, do not require the average relative gain of calibration; The general power that records is independent of SOP, and does not need " definitely " calibration directly to measure absolute transmission value; They can be acquired as and have a unknown constant factor.Subsequently on SOP, asking the standardization of average average power, as indicated above, eliminated the unknown factor.

Comprise at pick-up unit 22 under the situation of single detector 22A (for example Figure 1B), by calculating the mean value of all power in first and second power packages, and with each power divided by described average power, obtaining first and second groups of standardization power, such as hereinafter detailed description.

Figure 1B shows a kind of PMD surveying instrument that is fit to obtain by means of the standardization power that obtains in this way DGD or PMD.Surveying instrument is similar shown in the surveying instrument of PMD shown in Figure 1B and Fig. 1 C, but surveying instrument shown in Figure 1B has omitted coupling mechanism 21 and detecting device B 22C.Data processor 34 will use different standardization equatioies simply.

(see Fig. 1 H) under the situation of having used polarimeter 45, the different polarized component of several (normally three) of the light that comes out from FUT 18 can be according to the design of polarimeter, simultaneously or side by side be not measured to.

The single-ended surveying instrument that it should be noted that Fig. 2 can be adapted to be, it analyzer and pick-up unit 44 in use polarimeter 45.

In the both-end PMD based on polarized broadband light sources shown in Fig. 1 I measured, tunable optic filter 27 was used to the selective light wavelength.This tunable optic filter can be placed in polarizer 20A (Fig. 1 I) afterwards or before.It should be noted that this tunable optic filter must be the wave filter to polarization insensitive, and can be in the different selection of time different wavelength of this tunable optic filter.

In above-mentioned embodiment arbitrary, input SOP controller 14A and output SOP controller 14B operate by this way: the given SOP (it can be any SOP on the Poincare sphere) of the light that receives for the input end at it, the SOP of light that leaves its output terminal (no matter to input SOP 14A or output SOP 14B) will be any one other SOP among a plurality of equally distributed substantially SOP on the Poincare sphere, no matter this distribution is at random or definite.Typically, the quantity of input and output polarization state is about 100 to 100,000, but this quantity can be any feasible quantity that can reasonably cover Poincare sphere.Yet, for input and output SOP both, also can use one.The distribution that it should be noted that these SOP needs not be and can not be veritably at random usually.Therefore when using stochastic distribution really for convenience, " pseudorandom " may be more suitable term---because it implements easier also more cheap (in measuring process, the latter always is easy to be subjected to the influence of moving of FUT 18 in any case) than even SOP point lattice.

Detection system device 22, no matter be that single detector, pair of detectors, a wave filter add detecting device or a detector array, and the sampling or the sampling and ask averaging circuit unit 32, just like used in the normal business power meter known to a person of ordinary skill in the art.

Advantageously, control module 30B can be discrete computing machine.Yet, it should be noted that single computing machine can carry out the function of data processor 34 and control module 30B.

Can make various other modifications within the scope of the invention to above-mentioned embodiment.For example, tunable modulated light source 12, input SOP controller 14A and analysis and pick-up

unit

14B, 20 and 22 can make the modulated light source that enters FUT 18 have the different polarization state with other, and can analyze the final signal of the far-end that leaves FUT 18 or the device of power.

The polarimeter that uses in the instrument shown in Fig. 1 H (some beam splitters typically, it has parallel three or four analyzers and OPTICAL SENSORS) the approximate more than one polarized component of measuring this signal or power simultaneously, but other similar configurations also are feasible.Alternatively, I-SOP controller 14A can send into three or more predetermined input light SOP, for example have Mueller well known in the art (Mueller set) is set, and polarimeter can be used as analyzer and pick-up unit, shown in Fig. 1 G.

It should be noted that each group is not limited to a pair of modulating light pulse or a pair of modulating light pulse series.In fact, can use three or more different tight alternate wavelength by each group power, rather than MIN two tight alternate wavelength X _LAnd λ _U

Yet, also it should be noted that for the overall PMD of both-end and measure, if can know the rough PMD value of this FUT, often do not require a pair of above modulating light pulse and a pair of above light pulse.Otherwise described at automatic prescan as mentioned, a pair of above modulating light pulse or a pair of above light pulse series can be used to this acquisition process.

Also it should be noted that can be by obtaining asking on average on the input and output SOP at random in a large number of given constant mid point wavelength to the single DGD at fix midway point wavelength place, and described given constant mid point wavelength has two tight alternate wavelength.Therefore, by measuring the many single DGD at the different mid point wavelength place in the setted wavelength interval, also can be given in the DGD in the setted wavelength interval with wavelength variations, thus, average on all or the most of DGD by the different wave length place in this setted wavelength interval, can further calculate average DGD and RMS DGD.Perhaps, also can---it obtains by averaging---calculate RMS DGD on wavelength and/or SOP by mean square deviation.

It will be appreciated that also the mid point wavelength is defined as the mean value of two tight alternate wavelength, and be particularly useful for help to describe a basic wavelength under embodiment.Be not all to need the mid point wavelength definitely in the everywhere of this calculating, and the practical laser wavelength also not " setting " at mid point wavelength place.Only need to know stride---promptly be used in accumulation PMD arbitrary in calculating poor to wavelength, no matter the mid point wavelength is how, though the mid point wavelength be at random with the unknown.(when each group use more than one wavelength to the time, as mentioned above, the notion of introducing " centre wavelength " is useful as the wavelength " label " corresponding to particular group.This will further discuss hereinafter.)

Though the aforesaid operations method changes its mid point wavelength at each SOP, this is not an essential feature of the present invention.---in order to obtain the best possibility mean value (according to the definition of PMD) of DGD---can obtain good measurement result although by covering big range of wavelengths, with constant centre wavelength, or or even, not only with constant input and output SOP but also with the constant centre wavelength of a band presetted wavelength stride (or difference on the frequency), PMD of the present invention measures not work with setovering, and can provide acceptable PMD to measure.

Single-ended overall PMD measures

As indicated above, if will measure DGD/PMD from the end of FUT 18, so analyzer and detecting unit 44 and simulation and digital signal processing unit 40 can with the input optical controller 42, and with the both-end embodiment in be used to carry out the single control module 30 of the control function of

control module

30A and 30B, be positioned at the near-end of FUT 18 together.And because these parts are positioned at a place, so specific features can be combined, their member can be modified in due course.Now with reference to Fig. 2 to 2G single-ended surveying instrument configuration is described corresponding to Fig. 1 to 1G (double-end measurement instrument configuration).

Like this, Fig. 2 shows a kind of tunable single-ended overall PMD based on OTDR and measures facility, the double-end measurement instrument class of this facility and Fig. 1 seemingly, but in this facility, input optical controller device 42 and analysis and pick-up unit 44 are positioned at the near-end of FUT 18 jointly and share back reflective extraction apparatus 52, and back reflective extraction apparatus 52 will import I-SOP controller 14A via connector 16 and output A-SOP controller 14B is connected to FUT18.Back reflective extraction apparatus 52 is two-way, because it is transported to FUT 18 with light from I-SOP controller 14A, and back reflective light is transported to A-SOP controller 14B from FUT 18.Tunable light-pulse generator 12 is connected to the I-SOP controller 14A by PMF 29A in Fig. 1.

The optical fiber patch cords that has PC (FC/PC or FC/UPC) connector or optical fiber pigtail formula mirror 50 is connected to the far-end of FUT 18, forms a local reverberator with the far-end at FUT 18.In fact, can use the reverberator of any kind, if it can reflect back into the far-end of light from FUT 18 surveying instrument.

Compared to Figure 1 another change is, instrument shown in Figure 2 has single control module 30, and it is controlled tunable light-pulse generator 12, two

SOP controller

14A and 14B, samples and asks averaging unit 32 and data processor 34.In addition, the member of measuring unit shown in Figure 2 is similar or identical with the member of surveying instrument shown in Figure 1, and running in a similar fashion.Yet signal processor must be suitable for tackling the following fact: when the light that is extracted has comprised from passing through of light source 12 at least a portion length of FUT 18 and followed the light of also being got back to the back reflective extraction apparatus by back reflective through identical path.

Should note, above the term of mentioning in the linguistic context that single-ended overall PMD measures " tunable OTDR " is not limited to the commercial type OTDR of global function, and being meant a kind of like this facility: it can be provided for sending into the optical pulse of optical fiber, and subsequently can be only survey with execution time door (time-gate) average to those following pulses: with the corresponding pulse of reflection corresponding to the specific time delay distance of optical fiber end (promptly corresponding to).However, the use of OTDR allows the FUT end to be identified, and allows FUT length measured, makes that thus the time gate window is correctly selected.

It should be noted that various modifications and the alternative major part described referring to figs. 1 through the double-end measurement instrument of 1H can be applied in the single-ended surveying instrument shown in Figure 2.Now with reference to Fig. 2 B to 2G so amended single-ended surveying instrument configuration is described briefly.

In instrument shown in Fig. 2 B, input optical controller 42 and analyzer and detecting unit 44 shared same polarization Discr. (polarizer) 20A and I/O-SOP controllers 14, polarization Discr. 20A and I/O-SOP controller 14 all are two-way, they transport input light via connector 16 towards FUT 18, and the light that returns from FUT 18 along back reflective in the other direction.Thereby I/O-SOP controller 14 has made up the function of discrete I-SOP 14A and A-SOP 14B, but wherein, for the light that passes I/O-SOP controller 14 along one of both direction, disturbing partially must height correlation.The back reflective extraction apparatus comprises circulator (circulator)/coupling mechanism 52A, and it is connected to light source 12 by PMF 29A, and is connected to the input end of polarization Discr. (polarizer) 20A by the 2nd PMF 29B.Circulator/coupling mechanism 52A is transported to detection system with back reflective light, and this detection system is illustrated as single detector 22A in Fig. 2 B.The output terminal of polarization Discr. (polarizer) 20A is connected to the input end of two-way I/O-SOP controller by conventional fiber.Identical among other members and Fig. 2.

Being aligned in the factory of

PMF

29A and 29B is determined by this way: all luminous powers from tunable pulsed laser source 12 all are maintained at one of them axle last (routinely on " slowly " axle) of two axles of

optical fiber

29A and 29B substantially.Because being polarization, circulator/coupling mechanism 52A keeps, so this aligning continues to the attachment point of itself and PBS or polarizer.Each end at

PMF

29A and 29B is attached in the process of relevant member, regulates the azimuthal orientation of this PMF, to transfer to FUT 18 with guaranteeing the light pulse maximum.

In use, in instrument shown in Fig. 2 G, input light from optical controller 42 is admitted to FUT 18 via the joints of optical fibre 16, and the back reflective light that is caused by any local reflection (such as the Fresnel reflection from the far-end 50 of FUT 18) returns analyzer and pick-up unit 44 via the joints of optical fibre 16, enters I/O-SOP controller 14 on opposite direction.The SOP of this light is by SOP controller (or scrambler) 14 conversion, and this light is polarized Discr. 20 thereupon---specifically being PBS---resolves into two components of the SOP with quadrature, and 0 degree and 90 is spent the linear SOP of relative orientations typically.The first detecting device 22B is connected to one of two outputs of PBS 20, and to receive one of these quadrature components, back reflective extraction apparatus 52 (for example circulator/coupling mechanism) is connected to another output terminal (with respect to the back reflective light from FUT 18).The second detecting device 22C and then be connected in the back reflective extraction apparatus 52 output terminal of transmission from the light of PBS 20 is to receive another quadrature component.In case,,, come the power sum that is detected of self-detector 22B and 22C and total back reflective power (S as hereinafter describing so that relative detector efficiency, wavelength dependency, circulator loss or the like are taken into account through suitably calibration ₀) proportional.Back reflective light can be similar to side by side and be detected by detecting device 22B and 22C.

In instrument shown in Fig. 2 C, input optical controller device 42 comprises tunable light-pulse generator 12, and with analyzer and pick-up unit 44 shared back reflective extraction apparatuss, polarizer 20A and I/O SOP control device 14.The back reflective extraction apparatus is illustrated as circulator/coupling mechanism 52A.As before, input light from optical controller device 42 is admitted to FUT 18 via the joints of optical fibre 16, the back reflective light that any local reflection (such as Fresnel reflection) of the far-end 50 by FUT 18 reflect returns to be analyzed and detection optical controller device 44, and along entering I/O-SOP controller 14 in the other direction, this light returns polarizer 20A thereupon.Detecting device 22A is connected to the output terminal of circulator/coupling mechanism 52.

In instrument shown in Fig. 2 D, the back reflective light that any local of the far-end 50 by FUT 18 reflects is along returning I/O-SOP controller 14 in the other direction, and this light returns polarizer 20A thereupon, is coupled device 21 then and is divided into two

parts.Detecting device

22B and 22C are connected to two output terminals of coupling mechanism 21, to produce two power that are repeated to measure.

It should be noted that and always be not necessary to detect back reflective light simultaneously with two detecting device 22B and 22C.Also can be at different slightly time detecting back reflective light.

Also it should be noted that and also can use a detecting device with an optical switch 23.In this case, two detecting

device

22B and 22C can be replaced by a detecting device 22A and add an optical switch 23 (Fig. 2 E and 2F).This optical switch is used to the back reflective light from different light paths, and---it is from circulator (or coupling mechanism) 52A or PBS 20C (Fig. 2 F) or coupling mechanism 21 (Fig. 2 E)---is transformed into same detecting device, and be detected at different time from the back reflective light of different light paths thus.

Also should note, these the configuration in---in design and the design among Fig. 2 G among Fig. 2 (B, C and D)---based on PBS 20C based on polarizer 20A also can pass polarizer or PBS at light from tunable optical source before, obtain polarized light by this incident SOP from tunable optical source light is regulated.That is to say, if tunable (pulse) light source does not have polarization well, or different light SOP is arranged under the different wave length, do not require any additional polarizer, but still require between tunable (pulse) light source 12 and circulator/coupling mechanism 52A, to insert additional Polarization Controller.For this situation, 29A and 29B are preferably replaced by SMF.

At control module 30 under the control of---it also controls tunable laser source 20---, the sampling and ask averaging circuit 32 to use internal simulation-digital quantizer to come the corresponding electric signal that changes along with the time of coming self-detector 22 is sampled in a known way, to obtain corresponding electroresponse signal, corresponding electroresponse pulse signal can be sampled and ask average then, think that specific light spike train provides the average response pulse, and provide the back reflective luminous power for this series, on average provide the back reflective of described series luminous power by described average response pulse is asked on the most of the time of its duration, obtain a plurality of smooth back reflective power at last.The pre-filtering that this average " time " window (or " time gate ") can depend on sampling and ask average electron device.The average power of gained is used for drawing DGD or PMD value by data processor 34, promptly from the far-end of FUT 18 or Differential Group Delay of any other connector (DGD) or polarization mode dispersion (PMD).It will be appreciated that and use usual conversion, convert time delay to distance according to the refractive index of optical fiber, to obtain the length of optical fiber.

Except controlling sampling and asking averaging circuit 32, control module 30 also to control the wavelength of tunable pulsed laser source 12 and the I/O-SOP that is selected by I/O-SOP controller 14.More specifically, for each setting k of I/O-SOP controller 14, control module 30 makes respectively at least one pair of tight each other alternate wavelength X _L ^(k)And λ _U ^(k)The back reflective luminous power is measured at the place.This is defined as the mean value of the actual wavelength of this light pulse series to the mid point wavelength of light pulse series, promptly

λ_{k} = (λ_{L}^{(k)} + λ_{U}^{(k)}) / 2 .

(understand with being convenient to for convenience, mark L and U refer to respect to the mid point wavelength X _k" low " and " height " wavelength).

It will be appreciated that when this group comprises a pair of or a pair of above light pulse series, as defined above the mid point wavelength in fact in this group each to all being different.

A pair of or a pair of above wavelength in group also can be used to measure the power from the back reflective of the local reflection of FUT far-end, be FUT 18 extraction PMD values then.Yet, measure for single-ended PMD and can use a pair of above wavelength, unless approximately be worth (seeing hereinafter discussion for details) for automatic prescan obtains PMD about automatic prescan.The one wavelength correspondence of optimizing can satisfy

, wherein

v_{L}^{(k)} - v_{U}^{(k)} = δv,

And v _L ^(k)And v _U ^(k)Under v=c/ λ corresponding to wavelength to λ _L ^(k)And λ _U ^(k), wherein c is the light velocity in the vacuum.

It will be appreciated that also centre wavelength only is conceptual definition, only defines in order to be convenient to describe when a group comprises two above wavelength.Only comprise under the situation of two wavelength a group, centre wavelength is equivalent to certainly " mid point wavelength " defined above.Be not all to need centre wavelength Anywhere, and do not need accurately to make group to concentrate yet,, and do not need optical maser wavelength is arranged on central wavelength because this target's center's wavelength is defined as the mid point wavelength around certain target's center's wavelength what calculate.Only need the wavelength associated stride---promptly any wavelength that in accumulation PMD calculates, uses between poor---information, and no matter centre wavelength how.

I/O-SOP controller 14 is provided with different I-SOP and A-SOP with pseudo-random fashion so that the point of the SOP on the Poincare sphere of those ordinary representations can be evenly distributed on the surface of described ball, though this distribution be at random or some lattice uniformly.

Before the single-ended overall PMD measuring process of describing in more detail based on tunable OTDR,,, should notice that such theory is not restrictive with the Theory of Interpretation basis in order to promote understanding to this operation.

The various modifications of single-ended PMD measurement mechanism

The present invention has included the various adjustment to single-ended overall PMD surveying instrument shown in Figure 2.For example, in tunable light-pulse generator device 12, PMF 29A can be replaced by polarization adjuster 14 (seeing Figure 10 A), the input end that it is connected respectively to tunable pulsed laser source 12 and is connected to back reflective extraction apparatus 52 by non-polarization maintaining optical fibre.

If being polarization, the light path between the input end of the output terminal of tunable light-pulse generator device 12 and polarization Discr. 20 (for example PBS among Fig. 2) keeps, polarization keeps circulator 52 so, for example in Fig. 2, can be replaced by polarization and keep coupling mechanism (for example 50/50 coupling mechanism).Yet circulator has superiority, because it has provided between the dynamic area of about 3dB than 50/50 coupling mechanism more.

Can expect that also polarization Discr. 20 can be polarizer and coupling mechanism, shown in Fig. 2 B.In this case, detecting device B 22C will be connected to coupling mechanism 21, to receive unpolarized relevant back reflective light.

Keep if the light path between the input end of the output terminal of tunable pulsed laser source 12 and polarization Discr. 20 is not a polarization,---to be coupling mechanism or circulator 52---just need not be the back reflective extraction apparatus that polarization keeps so.

The patch cord that has FC/PC (or FC/UPC) connector or optical fiber pigtail formula mirror can be used to be connected the FUT far-end, is used to measure local reflection from the overall PMD of this FUT with manufacturing.

Preferably long from the light pulse length of tunable OTDR or duration, for example more than 1 to the 20 μ s, but also can use short pulse length or duration.

Though these modifications can be implemented discretely, the embodiment of the present invention shown in Fig. 2,2 (B-G) have comprised several such modifications.Particularly, light path between tunable pulsed laser source 12 and the I/O-SOP controller 14 is not that polarization keeps, the PMF 29A and the 29B that are Fig. 2 are replaced by the polarization state regulator, this polarization state regulator based on single-mode fiber (for example passes through, Corning, Inc. member (such as circulator and polarization beam apparatus 20) the non-PMF optical fiber of Ying Xiao SMF-28 by name) is connected, and the polarization state regulator makes the pulsed laser light maximizes power through I/O-SOP controller 14 then.

Replace the PBS 20 among Fig. 2 G, polarization Discr. 20 comprises polarizer 20A and coupling mechanism 21 combinations (Fig. 2 C), for 50/50 coupling mechanism situation, is cost to lose between the dynamic area that is similar to 3dB.The second detecting device B 22C (Fig. 2 G) is connected on one of them arm of coupling mechanism 21, is used for processing to detect a part of back reflective light, thereby derives total back reflective power of these pulses.

Those of ordinary skills need not excessive experiment just can make above-described process be suitable for being used for calibrating the relative sensitivity of these two detecting device A and B (22B and 22C), comprise owing to getting involved the loss that circulator or coupling mechanism or the like cause, so that be applied to the single-ended overall PMD surveying instrument of Fig. 2 G.It will be appreciated that in the embodiment of Fig. 2 C, do not require the average relative gain of calibration; The general power that records is independent of SOP, and does not need " definitely " calibration directly to measure absolute transmission value; Absolute transmission value can be acquired as and have the unknown constant factor.Subsequently on SOP, asking the standardization of average average power, as indicated above, eliminated the unknown factor.

Can imagine, comprise at pick-up unit 22 under the situation of single detector 22A (Fig. 2 B), by calculating the mean value of all power in first and second power packages, and with each power divided by described average power, to obtain first and second groups of standardization power, as hereinafter describing in detail.

Fig. 2 B has demonstrated a kind of single-ended PMD measurement that is fit to obtain by means of the standardization power that obtains in this way PMD.Single-ended overall PMD shown in Fig. 2 B measures with similar shown in Fig. 2 C, but has omitted coupling mechanism 21 and detecting device B 22C.Data processor 34 will use different standardization equatioies simply.

In above-mentioned arbitrary embodiment, I/O-SOP controller 14 operates in the following manner: the given SOP (it can be any SOP on the Poincare sphere) of the light that receives for the input end at it, the SOP of light that leaves its output terminal will be any among a plurality of equally distributed substantially SOP on the Poincare sphere, no matter this distribution is at random or definite.Typically, the output polarization amount of state is about 100 to 500, but this quantity can be any feasible quantity.Yet, also can use an I/O-SOP controller (rather than both-end PMD shown in Figure 1 measures two used SOP controllers).The distribution that it should be noted that these SOP needs not be and can not be veritably at random usually; Therefore when using stochastic distribution really for convenience, " pseudorandom " may be more suitable term---because it implements easier also more cheap than even SOP point lattice.

No matter detector means 22---be single detector or pair of detectors, and the sampling and ask averaging circuit unit 32, just like used in the normal business power meter known to a person of ordinary skill in the art.

Polarization Discr. 20 comprise PBS 20C or comprise polarizer 20A and the situation of the combination of coupling mechanism 21 under, situation for 50/50 coupling mechanism, the cost of losing between about 3dB dynamic area will be arranged, wherein the second detecting device 22C is connected on one of them arm of coupling mechanism 21, to detect a part of light, be used for handling, thereby derive total optical power, yet so the power that reduces may not be crucial to this measurement.

Advantageously, control module 30 can be discrete computing machine.Yet, it should be noted that single computing machine can carry out the function of data processor 34 and control module 30.

Single-ended accumulation PMD measures

Polarization-Sensitive optical time domain reflectometer (POTDR) shown in Figure 3 comprises all by tunable light-pulse generator device 12, the two-way Polarization Control apparatus 14 (being called as I/O SOP control device easily) of control module 30 controls, samples and ask averaging unit 32 and data processor means 34, and comprising pick-up unit 22, this pick-up unit comprises the first detecting device A 22B and the second detecting device B 22C.Tunable light-pulse generator device 12 is coupled to polarization maintaining optical fibre (PMF) 29A, be used for producing some are sent to tested optical fiber (FUT) 18 from connector 16 via I/O polarization state (I/O-SOP) control device 14 light pulse, and I/O polarization state (I/O-SOP) control device also as explained below, will receive corresponding back reflective light from FUT 18 via connector 16.

Input optical controller device 42 and analyzer and pick-up unit 44 comprise: the back reflective light extractor---and specifically be the polarization maintenance circulator 52 among Fig. 3; Polarization Discr. (PD) device 20---specifically be the polarization beam apparatus (PBS) among Fig. 3; And input and output SOP controller (or scrambler) 14.Circulator 52 is coupled to the input end of PBS 20 by the 2nd PMF 29B, so that keep polarization from the light path of tunable laser source 12 to PBS 20.Preferably, utilize single-mode fiber that PBS 20 is coupled to I/O-SOP controller (or scrambler) 14.

Being aligned in the factory of

PMF

29A and 29B is determined by this way: all luminous powers from tunable pulsed laser source 12 one of two axles of all being maintained at

optical fiber

29A and 29B are gone up (being maintained at routinely on " slowly " axle) substantially.Keep because circulator 52 is polarizations, so this aligning is held, up to the far-end of PMF 29B, it is attached to till that of PBS 20.Each end at

PMF

29A and 29B is attached in the process of relevant member, and the azimuthal orientation of this PMF 29A/B is conditioned, to transfer to FUT 18 with guaranteeing the light pulse maximum.

By reflecting the back reflective light that is caused, along entering I/O-SOP controller 14 in the other direction from the Rayleigh scattering light of FUT 18 and in some cases discrete (Fresnel).The SOP of this light is transformed by SOP scrambler 14, and this light is resolved into two components of the SOP with quadrature by PBS 20 thereupon, typically the linear SOP of 0 degree and 90 degree relative orientations.The first detecting device 22C is connected to one of two output terminals of PBS 20, and receiving in these quadrature components, circulator 52 is connected to another output terminal (with respect to from the back reflective light of FUT 18).Next, the second detecting device 22B and then be connected in the circulator 52 output terminal of transmission from the light of PBS 20 is to receive another quadrature component.In case,,, come the power sum that is detected of self-

detector

22B and 22C and total back reflective power (S as hereinafter describing so that relative detector efficiency, wavelength dependency, circulator loss or the like are taken into account through suitably calibration ₀) proportional.

At control module 30 under the control of---it also controls tunable laser source 12---, sampling and ask averaging circuit 32 to use internal simulation-digital quantizer that the electric signal of the correspondence that changes along with the time of coming self-

detector

22B and 22C detectable signal is sampled in a known way, to obtain corresponding surge response signal, then to sampling corresponding to the shock response signal of specific light spike train and asking average, to produce at this serial OTDR track.The OTDR track of gained is used for drawing accumulation PMD curve PMD (z) by data processor 34, that is, and and along with FUT 18 leaves the distance z of FUT 18 near-ends on along the line and the polarization mode dispersion (PMD) that changes.The near-end of described FUT18 is that end that is coupled to analyzer and pick-up unit 44.It will be appreciated that using usual conversion, convert time delay to distance according to optical fibre refractivity.

Except control sampling and ask averaging circuit 32, control module 30 also to control the wavelength of tunable pulsed laser source 12 and I/O-SOP controller 14 and select I-SOP and A-SOP right.More specifically, for the k of each setting of I/O-SOP controller 14, control module 30 makes respectively at least one pair of tight alternate wavelength X each other _L ^(k)And λ _U ^(k)The back reflective luminous power is measured at the place.This is defined as the mean value of the actual wavelength of this light pulse series to the mid point wavelength of light pulse series, promptly

λ_{k} = (λ_{L}^{(k)} + λ_{U}^{(k)}) / 2 .

It will be appreciated that when this group comprises a pair of above light pulse series, as defined above the mid point wavelength in fact in this group each to all being different.Must be appreciated that also centre wavelength only is conceptual definition, the enforcement under the only basic for convenience of description a pair of wavelength defines.Be not all to need centre wavelength Anywhere, and do not need accurately to make this that wavelength is concentrated on certain target's center's wavelength, because this target's center's wavelength is defined as the right mean value of actual wavelength yet what calculate.Do not need optical maser wavelength is arranged on central wavelength yet.Only need be about the wavelength stride---promptly any wavelength that in accumulation PMD calculates, uses between poor---information, no matter centre wavelength is how, even if also its be at random or unknown.

It is right that I/O-SOP controller 14 is provided with different (I-SOP and A-SOP) with pseudo-random fashion, so that the point of those ordinary representations SOP corresponding with this each right composition is evenly distributed on the surface of described Poincare sphere, and though this distribution be at random or some lattice uniformly.

Before the running of describing POTDR in more detail,,, should notice that such theory is not restrictive with the Theory of Interpretation basis in order to promote understanding to this operation.

The various modifications of single-ended accumulation PMD measurement mechanism

The present invention has included the various modifications to embodiment shown in Figure 3.For example, in tunable light-pulse generator device 12, PMF 29A can be replaced by polarization adjuster 14 (seeing Figure 10 A), the input end that this polarization adjuster is connected respectively to tunable pulsed laser source 12 and is connected to back reflective extraction apparatus 52 by non-polarization maintaining optical fibre.

Keep if the light path between the input end of the output terminal of tunable light-pulse generator device 12 and polarization Discr. 20 (for example PBS among Fig. 2) is a polarization, the polarization among Fig. 3 keeps circulator 18 can be replaced by polarization and keeps coupling mechanism (for example 50/50 coupling mechanism) so.Yet, preferably should be circulator, because it has given than 50/50 coupling mechanism more approximately between the dynamic area of 3dB.

Keep if the light path between the input end of the output terminal of tunable light-pulse generator device 12 and polarization Discr. 20 (for example PBS among Fig. 2) is not a polarization,---to be coupling mechanism or circulator 52---need not be the back reflective extraction apparatus that polarization keeps so.

Though these modifications can be implemented discretely, embodiment of the present invention shown in Figure 3 have comprised several such modifications.Particularly, light path between tunable pulsed laser source 12 and the I/O-SOP controller 14 is not that polarization keeps, promptly, PMF 29A and the 29B of Fig. 3 are replaced by polarization state regulator 14, this polarization state regulator based on single-mode fiber (for example passes through, Corning, Inc. member (such as circulator 52 and polarization beam apparatus 20) the non-PMF optical fiber of Ying Xiao SMF-28 by name) is connected, and this polarization state regulator makes through I/O-SOP controller 14 and the pulsed laser light maximizes power that enters FUT18.

Replace PBS 20, polarization Discr. 20 can comprise the combination of polarizer 20A and coupling mechanism 21, shown in Fig. 3 B, for the situation of 50/50 coupling mechanism, is cost to lose between the dynamic area that is similar to 3dB.The first detecting device 26A is connected on one of them arm of coupling mechanism 20A, detecting a part of back reflective light, by handling, thereby derives total back reflective power of these pulses.

In the POTDR of Fig. 3, then can carry out and the similar process of above-mentioned embodiment as described in Figure 3, although do not require aforesaid relative sensitivity, comprise that the loss that is caused by the circulator of getting involved or coupling mechanism or the like calibrates to two detecting

device

22B and 22C.

Those of ordinary skills need not excessive experiment just can be made above and be suitable for using in the embodiment of Fig. 3 with reference to the described calibration process of the POTDR of figure 3.It will be appreciated that in the embodiment of Fig. 3 B, do not require the average relative gain of calibration; The general power that records is independent of SOP, and does not need " definitely " calibration directly to measure absolute transmission value; Absolute transmission value can be acquired as and have the unknown constant factor.Subsequently on SOP, asking the standardization of average average power, as indicated above, the unknown factor will be eliminated.

Can imagine, pick-up unit 22 can comprise single detector, by calculating the mean value of all the OTDR tracks in first and second groups of OTDR tracks, and with each OTDR track divided by described average power, to obtain first and second groups of standardization OTDR tracks, describe in detail as mentioned.

Fig. 3 A shows a kind of POTDR that is fit to obtain by means of the standardization OTDR track that obtains in this way PMD.Similar shown in POTDR shown in Fig. 3 A and Fig. 3 B, but omitted coupling mechanism 21 and detecting device B 22C.The different standardization equation that provides in the method for operating that data processor 34 will use above to be provided simply.

In above-mentioned arbitrary embodiment, I/O-SOP controller 14 operates in the following manner: for the given SOP (it can be any SOP on the Poincare sphere) of the light that receives at its input end, the SOP that leaves the light of its output terminal will be any among a plurality of equally distributed substantially SOP on the Poincare sphere, no matter this distribution is at random or definite.Typically, for obtaining high quality results, I-SOP and A-SOP quantity separately is about 100 to 200, but can be any feasible quantity.It should be noted that each distribution of I-SOP and A-SOP does not need, also usually not can, be veritably at random; Therefore when using stochastic distribution really for convenience, " pseudorandom " may be more suitable term---because it implements easier also more cheap than uniform I-SOP and A-SOP point lattice.

Although preferably use two detecting devices side by side to obtain the component of two crossed polarized lights, can expect that also two detecting devices of this of the embodiment of Fig. 3 and 3B can be replaced by a detecting device and add an optical switch.This optical switch is used to lead back to catoptrical two orthogonal polarization components (Fig. 3), perhaps one of (for example optionally) guiding from the output of polarizer and another directly from the output (Fig. 3 B) of coupling mechanism, enter same detecting device, thereby make two orthogonal polarization components of back reflective light, perhaps one from the output of polarizer and another directly from the output of coupling mechanism, can be detected successively by same detecting device.

By will be corresponding to two of this series in the OTDR track of detected different polarized component at least one divided by OTDR track sum corresponding to two of this series detected different polarized components, can obtain standardization OTDR track at this light pulse series.This substituting scheme can be used, no matter analyzer and detector cell comprise PBS or coupling mechanism.Can estimate, all be small to any modification of standardization and processing, and in those of ordinary skills' general knowledge scope.

Alternatively, utilize a this detecting device to add the configuration of an optical switch, can detect a polarized component and total optical power successively by same detecting device.The same before this, this optical switch is used to a polarized component and total reference light power are imported same detecting device, and, OTDR track by should series is divided by the OTDR track corresponding to general power of this series, with the standardization OTDR track that obtains corresponding to this specific light spike train.Also it should be noted that disadvantageously,, use to have an optical switch and connect same detecting device and replace two detecting devices, total Measuring Time is doubled compared with the embodiment of using two detecting devices.

Can expect that also rotatory polarization Discr. (PD) be no matter---being polarizer or PBS---can be used to obtain successively two quadrature components, for example via with polarization Discr. half-twist with from detecting P _xSwitch to and detect P _y, or from detecting P _ySwitch to and detect P _xNo matter detector means 22---be single detector or pair of detectors---and the sampling and ask averaging circuit unit 232, just like used among the normal business OTDR known to a person of ordinary skill in the art.

Advantageously, control module 30 can be a discrete computing machine.Yet, it should be noted that single computing machine can carry out the function of data processor 34 and control module 30.

Can make various modifications to above-mentioned embodiment within the scope of the invention.For example, tunable pulsed laser source 12 and I/O-SOP controller 14 can be replaced by some other can provide the different polarization state of the pulse that enters FUT 18, and can analyze final leave FUT 18 by Rayleigh scattering and/or the caused back reflective light signal of discrete reflectance.

Like this, can use polarimeter (some spectrometers, it has parallel three or four analyzers and OPTICAL SENSORS)---it measures the more than one polarized component of back reflective signal simultaneously---or some other configuration, depend on back reflective polarized state of light (SOP) so that arrive the power of OPTICAL SENSORS.

It should be noted that each group is not limited to a pair of light pulse series.In fact, advantageously, each the group track for obtaining under common SOP uses three or more different tight alternate wavelength, rather than MIN two tight alternate wavelength X _LAnd λ _USo (each group comprises 2N _λIndividual OTDR track, rather than four, in embodiment, be two groups of 2N with two OPTICAL SENSORS _λIndividual track, wherein N _λBe the quantity of one group of wavelength in the light pulse series).For example, under the situation of having used three tight alternate wavelength, the light pulse series that can be chosen in minimum and intermediate wave strong point is as a pair of, the light pulse series that is chosen in middle and the highest wavelength place is as second pair, thereby make wavelength stride between a pair of light pulse greater than another to the wavelength stride between the light pulse, perhaps go out several times greatly.

Because corresponding to three wavelength three wavelength combinations being arranged (is N _λ(N _λ-1)/2), so can obtain data corresponding to two remarkable different wavelength strides simultaneously in the following time: the described time is only measured 1.5 times of required time for carrying out a stride formula.Therefore, utilize each three wavelength of group (or more a plurality of) to operate, be proved to be highly favourable, because accumulation PMD value can increase (the overall PMD from 0 to this FUT) significantly along the length of FUT 16, thereby, use two, three or more different wavelength strides to make and can on all positions of this optical fiber, all keep gratifying relative accuracy (for example representing) with number percent.It will be appreciated that the spectrum that also can be chosen in minimum and the highest wavelength place is listed as the 3rd pair, it has than first pair and second pair of all big wavelength stride.Use only stride, provided a given degree of certainty anything but, for example ± 0.1ps, the distance that its value at PMD rises to 10ps has been represented a less number percent uncertainty, but in the PMD short distance of 0.2ps only for example, aspect the number percent of measuring accuracy and bad.In order to obtain less uncertainty of measurement, must select bigger wavelength stride for less PMD value.Thereby, obviously advantageously, implement the wherein alternate embodiment of two above wavelength of each group use.It does not change aforesaid setting of the present invention and principle, has but saved the time of whole measurement and processing.

Though above-mentioned embodiment has changed the centre wavelength of each SOP, this is not an essential feature of the present invention.---in order to obtain the best possibility mean value (according to the definition of PMD) of DGD---can obtain good measurement result although by covering big range of wavelengths, but with constant centre wavelength, POTDR of the present invention is the work of bias ground not, and can provide acceptable PMD to measure.

Basic theory, data processing and computing method

Though the applicant does not wish to be bound by theory, and following basic theory is provided, to promote understanding to various embodiments of the present invention.

Calculating to DGD or rms DGD (being PMD)---it disturbs the PMD measuring principle of inclined to one side analysis (SSA) method based on input and output polarization state at random---utilizes the measure theory of the relevant PMD of prior art, comprise Poincare sphere analysis (PSA) and expansion interferometric method (GINTY), through suitably adjusting, draw equation given below.Be applied to concrete theory and the international patent application No.PTC/CA2006/001610 and the above-mentioned U.S. part continuation application No.11/727 of various aspects of the present invention, the theory of describing in 759 is closely related, and the full content of these two documents is included this instructions by reference in.

In whole instructions, used wavelength X---wherein λ is a light wavelength in a vacuum---and optical frequency v, but they have the known λ of relation=c/v certainly.Though the use of optical frequency is more " nature " in this theory, in practice, for tight alternate wavelength, wavelength can be used, should be understood that suitable conversion factor is applied to the equation in this instructions.

Should recall, PMD is the statistics RMS value of Differential Group Delay DGD (λ), it is estimated in the following manner, promptly on big range of wavelengths or---to be on the two ideally---in a period of time average and assess, so that DGD random case as much as possible is observed, to obtain its RMS value.

Basic theory

The SOP of I/O at random that is used for the PMD measurement disturbs inclined to one side analysis

At these chapters and sections, we will describe the basic theory of " polarization state of input and output at random that is used for polarization mode dispersion measurement is disturbed inclined to one side analysis (SSA) method ", with and application in the PMD measurement of being undertaken by the both ends or one end of inserting FUT.These three main application are: (1) " is used for determining the DGD of optical link and both-end PMD measuring method and the facility of PMD " (abbreviating " both-end PMD measurement " as); (2) " use the single-ended overall PMD of tunable OTDR measure with and the method for definite PMD " (abbreviating " single-ended overall PMD measures " as); (3) " Polarization-Sensitive optical time domain reflectometer (POTDR) with and detect the method for the accumulation PMD change along with fiber lengths " (abbreviating " single-ended accumulation PMD measurement " as).Method of operating, data processing and computing method during these are used will be described in detail at following chapters and sections.

If tunable laser and Polarization Controller are used to send and be controlled at the input light of the end incident of FUT, and polarization state analyzer and power meter are used to measure two tight alternate optical frequency v from this FUT (from the same end or the other end of this FUT) _UAnd v _LUnder the power of a large amount of I/O polarization states of K, described optical frequency v _UAnd v _LRound given midpoint frequency v _Mid, the power of a large amount of I/O polarization states of described K promptly comprises a large amount of " SOP to " (I-SOP _k, A-SOP _k), each SOP is to not only relating to input SOP but also relating to the polaroid analyze axle of " being seen " by received light.I-SOP and A-SOP value all should be selected with random fashion, so that the point of the SOP on the expression Poincare sphere is evenly distributed on the surface of described ball routinely, no matter this distribution is at random or uniform some lattice.Have been found that by on the enough big even distribution number K of described " SOP to ", averaging, at v _UAnd v _LLocate mean square deviation and midpoint frequency v between the observed standardization power _Mid(v _Mid=(v _U+ v _L)/2) have simple relation between the DGD that locates, it is all effective that this pass ties up under all scenario the actual FUT for any kind, no matter its degree at random or its polarization coupled rate are how, comprising this egregious cases of PMF optical fiber, promptly

DGD (v) = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{{< ΔT {(v)}^{2} >}_{SOP}}) - - - (1)

Wherein＜ _SOPThe mean value of representative on K SOP, δ v=(v _U-v _L) be " frequency step ", α _DsBe theoretical constant, it depends on the configuration of measuring on arranging, i.e. two or an end measurement configuration.Δ T (v) is at v _UAnd v _LLocate the poor of the observed respectively standardization power of analyzing (i.e. transmission), its mean square deviation is

{< ΔT {(v)}^{2} >}_{SOP} = {< {(T_{U} - T_{L})}^{2} >}_{SOP} = \frac{1}{K} \underset{k}{Σ} {(T_{U}^{(k)} - T_{L}^{(k)})}^{2}

Wherein, for shown in Figure 1B, 2C and 3A based on the embodiment under the situation of a detecting device of polarizer, standardization power is

T_{L}^{(k)} = u_{o} \frac{P_{L}^{(k)}}{{< P_{L} >}_{SOP}}

T_{U}^{(k)} = u_{o} \frac{P_{U}^{(k)}}{{< P_{U} >}_{SOP}}

Wherein with reference to average u ₀Be theoretical constant, it depends on the configuration of measuring on arranging, and promptly both-end (Figure 1B) or single-ended (Fig. 2 C and 3A) measure configuration, and average power is defined as

{< P_{L} >}_{SOP} = \frac{1}{K} \underset{k}{Σ} P_{L}^{(k)} .

{< P_{U} >}_{SOP} = \frac{1}{K} \underset{k}{Σ} P_{U}^{(k)}

In addition, range of wavelengths for regulation, in a preferred embodiment of the invention, mean value shown in the formula (1) is preferably not only on many " SOP to " but also try to achieve on many mid point wavelength, preferred one group of closely alternate two pairing SOP of wavelength and its mid point wavelength will constantly change to the next one from one, with this tried to achieve rms DGD on the range of wavelengths of regulation (also promptly: PMD), be expressed as:

PMD = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{{< ΔT {(v)}^{2} >}_{SOP; λ}}) - - - (2)

Wherein＜ _{SOP; λ}Be not only to ask average on the SOP but also on wavelength, or average on the wavelength in the provision wavelengths interval.

Under the restriction of the enough little optical frequency poor (" frequency step ") between the closely alternate wavelength, it is as follows that equation (1) and (2) are reduced to simpler differential formulas

DGD (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{{< ΔT {(v)}^{2} >}_{SOP}} - - - (1 a)

PMD = \frac{α_{ds}}{πδv} \cdot \sqrt{{< ΔT {(v)}^{2} >}_{SOP; λ}} - - - (2 a)

DGD that draws from above-mentioned equation (1) and (2) and PMD are effectively for both-end and single-ended measurement configuration, and they represent the measured value between output port and the output port.For double-end measurement configuration, theoretical constant α _DsBe

α_{ds} = \sqrt{\frac{9}{2}} - - - (3)

And for single-ended measurement configuration, if use of the control of common (identical) polarization state controller (scrambler) as input and output light SOP, such as for Fig. 2,2C-G, theoretical constant α _DsBe

α_{ds} = \sqrt{\frac{15}{4}} - - - (4)

For different measurement configurations, with reference to average u ₀Also be different.For the double-end measurement configuration, with reference to average u ₀Be

u_{0} = \frac{1}{2} - - - (5)

And, for single-ended measurement configuration, if the incident polarization state (I-SOP) of light is parallel to the analyzer axle, for example in Fig. 2 C, with reference to average u ₀Be

u_{0} = \frac{2}{3} - - - (6)

Note, the relation in the equation (1) for double-end measurement configuration be DGD δ v＜ ³/ ₄Set up under the condition, for single-ended measurement configuration be DGD δ v＜ ¹/ ₂Set up under the condition, so just clarified the implication of " closely alternate wavelength ".

Should note, the DGD that calculates from equation (1) and (2) (v) and PMD respectively, be definite DGD and the PMD value of measuring between input connector (16A) at FUT and the out connector (16B), and they can not represent unidirectional (forward) DGD or the PMD of this FUT, for example, and for single-ended measurement configuration, DGD that records and PMD value are the round values of FUT, but, for double-end measurement configuration, unidirectional (forward) DGD or PMD that DGD that records that draws from equation (1) and (2) or PMD are this FUT.Measure configuration for single-ended PMD, require on the round PMD that records that equation (2) draws, to multiply by a round factor

(α_{rt} = \sqrt{\frac{3}{8}}),

To obtain unidirectional (forward) PMD value of FUT.

In each embodiment, obtaining of standardization power in fact is different, promptly passes through the suitable programming of data processor 34.The explanation of this theory is that the basic OPTICAL SENSORS embodiment for Figure 1B, 2C and 3A provides, in these embodiments, standardization on average power be necessary be again fully, suppose that general power is in that (I-SOP is A-SOP) to being stable when changing along with the time.It should be noted that the standardisation process at double-end measurement configuration (Figure 1B) and single-ended measurement configuration (Fig. 2 C and 3A) is very similar, but with reference to average (u ₀) (seeing equation (5) and (6)) be different.It should be noted that also for single-ended accumulation PMD and measure that the standardization power track that changes along with distance z (T (z)) is calculated.The detailed description of this standardisation process provides hereinafter.

It should be noted that equation (1) has produced is giving fix midway point wavelength---it is defined as being used in the mean wavelength of the specific tight alternate wavelength in this measurement---the DGD value of locating, and has provided the DGD that changes along with optical wavelength/frequency.Equation (2) has produced the PMD value at the range of wavelengths of regulation.This PMD is defined in DGD root mean square (rms) value that averages on the wavelength.

Both-end PMD measures

In this area used most available PMD measuring technique normally both-end PMD measure.The basic theory of the above-mentioned SSA of input and output at random method can be applied to both-end PMD and measure, and wherein test link can comprise or not comprise optical amplifier.When optical amplifier is used in the test link, will be mixed in the polarization coherent light of being sent into from the ASE light of amplifier, therefore, ASE and the light of being sent into all will be measured by OPTICAL SENSORS 22A (Figure 1B).

Below how we describe by inserting the two ends of FUT, our SSA basic theory is applied to having and do not have the both-end PMD measuring method that these two kinds of situations of optical amplifier all are suitable in the test link.

The DGD that does not have image intensifer in test link measures

If tunable laser source---it can pass through phase step type tuning (step tuning), or frequency sweeping, or frequency modulation (PFM), if or similar fashion is selected its optical frequency---perhaps use polarized broadband light sources, so available tunable optic filter is selected optical frequency (wavelength), and the input polarization controller is placed in the near-end of FUT, and polarization state analyzer---output polarization controller normally, polarizer (or PBS) and power meter (if combine with tunable optic filter---used is polarized broadband light sources, but not tunable laser source)---be placed in the other end of FUT, be used to measure two tight alternate optical frequency v from this tested optical fiber _UAnd v _LUnder the luminous power of K I/O polarization state, described optical frequency v _UAnd v _LRound given midpoint frequency v _Mid, the luminous power of a large amount of I/O polarization states of described K promptly comprises a large amount of " SOP to " (I-SOP _k, A-SOP _k), each SOP is to not only relating to input SOP but also relating to the analyzer axle of " being seen " by received light.I-SOP and A-SOP should be selected with pseudo-random fashion, are evenly distributed in substantially on the surface of described ball so that be illustrated in the point of the SOP on the Poincare sphere routinely, no matter this distribution is at random or the intimate lattice distribution of point uniformly.Have been found that by on the enough big even distribution number K of described " SOP to ", averaging, at midpoint frequency v _Mid(v _Mid=(v _U+ v _L)/2) DGD forward that locates can calculate by equation (1), for

DGD (v) = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{{< ΔT {(v)}^{2} >}_{SOP}}) - - - (7)

It should be noted that equation (7) has produced this FUT in unidirectional (forward) DGD value (being DGD) that given midpoint frequency (wavelength) is located.

As already mentioned, PMD is defined in DGD root mean square (rms) value (it should be noted that the DGD that asks average in time can provide rms DGD (rather than average DGD)) of asking average on the wavelength.Rms DGD (being PMD) on the range of wavelengths of regulation can calculate by equation (2) now, for

PMD = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{{< ΔT {(v)}^{2} >}_{SOP; λ}}) - - - (8)

What need attention once more is in equation (7) and (8), to measure configuration for both-end PMD, the necessary use

α_{ds} = \sqrt{\frac{9}{2}} .

Satisfy in " frequency step " and concern under the condition of DGD δ v＜3/4, so just clarified the implication of " tight alternate wavelength ".

Under the restriction of the enough little optical frequency poor (" frequency step ") between the closely alternate wavelength, it is as follows that equation (7) and (8) are reduced to simpler differential formulas

DGD (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{{< ΔT {(v)}^{2} >}_{SOP}} - - - (7 a)

PMD = \frac{α_{ds}}{πδv} \cdot \sqrt{{< Δ T^{2} >}_{SOP; λ}} - - - (8 a)

It should be noted that the basic theory in the direct adaptive type (1) of equation (7) and (8) energy and (2), to calculate DGD forward and the PMD of FUT.

The DGD that has amplifier in test link measures

In many rig-site utilization, optical amplifier (er-doped optical amplifier typically) has been inserted into this link.Being that FUT 18 can comprise at least one with various intervals (for example 60km) in FUT 18, also can be several, optical amplifier.When having optical amplifier, the power meter of far-end that is positioned at FUT 18 also may detect (unpolarized substantially) amplified spontaneous emission (ASE) light except the signal that detects optics generating apparatus emission.Spontaneous emission light is equivalent to that " dwindling " mean square deviation＜Δ T (v) by going with a factor ² _SOPValue, and the existence of ASE in the detection signal is taken into account, the described factor can be calculated with same raw data.This factor sigma may _r ²(v) be the relative variance of standardization power, be defined as

σ_{r}^{2} (v) = {(\frac{1}{σ_{20}})}^{2} [{< T {(v)}^{2} >}_{SOP} - {< T (v) >}_{SOP}^{2}] - - - (9)

Wherein be with reference to variance

σ_{20}^{2} = 1 / 12 .

＜T (v) ² _SOPWith＜T (v) 〉 _SOP ²Refer to v _UAnd v _LThe average of the standardization power at place (it should be noted that for the standardization power T that is averaged on the inclined to one side SOP disturbing at random of quantity sufficient (v),

{< T (v) >}_{SOP}^{2} = \frac{1}{4}) .

Then, by with mean square deviation divided by the relative variance in the equation (9), can obtain DGD forward (unidirectional) at given mid point wavelength place, for

DGD (v) = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{\frac{{< ΔT {(v)}^{2} >}_{SOP}}{σ_{r}^{2} (v)}}) - - - (10)

In addition, the DGD of rms forward (unidirectional) of the range of wavelengths of appointment can be expressed as

PMD = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{\frac{{< Δ T^{2} >}_{SOP; λ}}{σ_{r}^{2}}}) - - - (11)

Wherein, showing now average on the two of SOP and wavelength in the equation (10) average on the SOP, and the relative variance of standardization power is expressed as now

σ_{r}^{2} = {(\frac{1}{σ_{0}})}^{2} [{< T {(v)}^{2} >}_{SOP; λ} - {< T (v) >}_{SOP; λ}^{2}] - - - (12)

Under the restriction of little stride, equation (10) and (11) are reduced to differential formulas, for

DGD (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< ΔT {(v)}^{2} >}_{SOP}}{σ_{r}^{2} (v)}} - - - (10 a)

PMD = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< Δ T^{2} >}_{SOP; λ}}{σ_{r}^{2}}} - - - (11 a)

Should note, two transmitted powers of " if closely alternate wavelength " be equate and, for these " closely alternate wavelength ", spectrum decay from FUT is insignificant, the power at " closely alternate wavelength " that records so can directly apply to equation (10) and (11), promptly for measured power without any need for standardization (it should be noted that in this case,＜T (v) 〉 _SOP ²May no longer equal 1/4).This be because, under this situation, above-mentioned standardisation process can only produce a quilt take advantage of on the power that records in order to obtain " constant factor " of standardization power (between 0 and 1), but when calculating DGD and PMD by use equation (10) and (11), this constant " factor " is finally by cancellation, and definite identical " factor "---they are all directly calculated from the power that records because mean square deviation and relative variance are all taken advantage of on the two.In other words, if use equation (10) and (11), only require acquisition and the proportional relative power of standardization power to calculate DGD and PMD.

It should be noted that equation (10) and (11) can be applied to have in the test link or do not have the situation of amplifier " noise ".

In an alternative method of the present invention, can be by (v) value is asked root mean square or asked average, obtains the estimation to PMD (i.e. rms on an optical frequency interval or average DGD value) to all single DGD at the indicated different mid point wavelength place in equation (7) or (10) on an optical frequency interval.

Single-ended PMD measures

It is very important measuring technique that single-ended PMD measures for rig-site utilization.Above-mentioned SSA basic theory also can be applied to single-ended PMD and measure.Single-ended PMD measurement described herein is divided into two kinds of situations: first kind of situation is by analyzing the back reflective light from the far-end of FUT, to measure all overall PMD of this FUT; Second kind of situation is to measure the accumulation PMD that changes along with FUT length.These two kinds of situations all only need insert the end of FUT.

Single-ended overall PMD measures

Single-ended PMD measurement for using from the back reflective light of optical fiber far-end may usually relate to the measuring fiber of not being with optical amplifier.Below we describe that our basic SSA of application is theoretical to be come bringing in and carry out single-ended overall PMD and measure by only inserting one of FUT.

If a catoptron (such as the optical fiber pigtail mirror) is connected the far-end of this FUT, if and can ignore rayleigh backscattering, with any pseudo-discrete reflectance along this FUT (for example, from locating of any connector or joint), tunable OTDR can be replaced by tunable CW laser instrument (no pulse) and power meter, and this power meter is used to measure two tight alternate optical frequency v that reflective mirror reflected of FUT far-end _UAnd v _LUnder a large amount of (I-SOP of K _k, A-SOP _k) right luminous power, described optical frequency v _UAnd v _LRound given midpoint frequency v _Mid, described (I-SOP _k, A-SOP _k) to be the setting relevant with the polaroid analyze axle of importing SOP and " seen " by back reflective light (note, λ=c/v, wherein λ is a light wavelength in a vacuum.Though in this theory, use optical frequency more " nature ", in practice,, can use wavelength for tight alternate wavelength, should be understood that these equatioies that suitable conversion factor is applied to presenting herein.)。From above-mentioned basic PMD measure theory have been found that enough big equally distributed K described (I-SOP, A-SOP) on carry out mean time, at v _UAnd v _LLocate the mean square deviation between the observed standardization power (i.e. transmission), with its midpoint frequency v _c(v _c=(v _U+ v _L)/2) the round DGD that locates (v) exists the simple relation shown in equation (1), this relation is all effective for the actual FUT of any kind, no matter how are the degree at random of this FUT or its polarization coupled rate, wherein also comprise this egregious cases of PMF optical fiber, this relation is as follows

{DGD}_{RoundTrip} (v) = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{{< ΔT {(v)}^{2} >}_{SOP}}) - - - (12)

Wherein, measure theoretical constant value for single-ended round DGD

α_{ds} = \sqrt{\frac{15}{4}},

＜ _SOPRepresentative K (I-SOP, A-SOP) to last mean value, δ v=(v _U-v _L) be " frequency step ", Δ T is at v _UAnd v _LLocate the poor of the observed standardization power of difference.

This relation is for DGD _RoundTripSet up under the condition of δ v＜1/2, so just clarified the implication of " closely alternate wavelength ".

The round DGD that is drawn by equation (12) (v) is not DGD (doubling v) forward.Rms DGD from the range of wavelengths (the v) round DGD of Ti Quing _RMSNeither double.Yet for a kind of situation in back, when carry out mean time at wavelength or on the time, PMD value (statistical average) (being rmsDGD) (is rms DGD by a simple factor and round PMD _RoundTrip) relevant, this round factor

α_{rt} = \sqrt{3 / 8},

Be DGD _RMS=α _RtDGD _RoundTripRMSOr

PMD = \sqrt{3 / 8} \cdot {PMD}_{RoundTrip},

Wherein PMD is defined as root mean square (rms) value of DGD.

If it should be noted that the substituting definition of having used PMD---be the mean value of DGD, rather than this definition of RMS-DGD, have the different round factors so.

Typically, in order to measure overall PMD reliably, should use tunable OTDR.The pulse transmission that tunable OTDR will grow relatively is in FUT, and then, the OPTICAL SENSORS of at least one among this OTDR detects the back reflective luminous power of the local reflection of FUT far-end.

Being comprised in the output terminal of this instrument and the round DGD of the part of the FUT between the selected reflection obtains from equation (12) as previously mentioned, wherein at given (I-SOP, A-SOP) to observed power, now, for example be to ask average and obtain at a predetermined portions in this duration of pulse from the power of the pulse of selected reflection back reflective.

It should be noted that, by each response impulse is averaged, can obtain back reflective power defined above, therefore preferably on quite a few of its duration, for this single-ended PMD measuring technique, can apply long OTDR pulse (for example 1 to 20 μ s).

In addition, in a preferred embodiment of the invention, if overall total PMD is measured in expectation, so preferably pairing I-SOP, the pairing right group of wavelength of A-SOP and mid point wavelength should constantly change to the next one from one, equation (12) is average with shown in carrying out on this three, thereby obtains to come and go PMD, rather than a certain wave strong point is obtained a specific DGD.Round rms DGD on the range of wavelengths of regulation (promptly coming and going PMD) is expressed as:

{PMD}_{RoundTrip} = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{{< ΔT {(v)}^{2} >}_{SOP; λ}}) - - - (13)

In addition, by equation (13) be multiply by the above-mentioned round factor

α_{rt} = \sqrt{3 / 8},

Can obtain PMD value forward,

PMD＝α _rt·PMD _RoundTrip (14)

Under the restriction of the enough little optical frequency poor (" frequency step ") between the closely alternate wavelength, equation (12) and (13) are reduced to simpler differential formulas, and are as follows,

{DGD}_{RoundTrip} (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{{< ΔT {(v)}^{2} >}_{SOP}} - - - (12 a)

{PMD}_{RoundTrip} = \frac{α_{ds}}{πδv} \cdot \sqrt{{< ΔT {(v)}^{2} >}_{SOP; λ}} - - - (13 a)

The PMD that records based on equation (13) has short this advantage of Measuring Time.Yet, also can be from the DGD of a lot of different mid point wavelength that records _RoundTrip(v),, obtain rms DGD by equation (12) or (12a) root mean square on the range of wavelengths of regulation or average _RoundTripOr average DGD _RoundTrip, for example

rms {DGD}_{RoundTrip} = \sqrt{{< {DGD}_{RoundTrip}^{2} >}_{λ}}

With average DGD _RoundTrip=＜DGD _RoundTrip _λThen, by simply respectively to rms DGD _RoundTripWith average DGD _RoundTripMultiply by the round factor

With 2/ π, obtain rms DGD and average DGD forward.

Single-ended accumulation PMD measures

By each distance (z) on the FUT length is analyzed rayleigh backscattering light, equation (12) and (13 of describing at single-ended overall PMD measurement) can be used to measure the single-ended accumulation PMD that changes along with distance z above.Like this, just be necessary to use short light pulse, for example from the light pulse of tunable OTDR.Note, use too short light pulse to limit and to survey FUT length, but that oversize pulse may not be handled the bat of optical fiber is long.

In fact, if used very short light pulse, so, OTDR " track ", or the back reflective power that changes along with distance z is identical, has been repeated unlimited number of times just like above-mentioned single-ended overall PMD measurement, and end reflector displacement increment dz between each time measured.If pulse is very short, and ignore " coherent noise ", just can obtain and the identical result of equation (12) always add this fact of OTDR track to, except this result be in a step, be retrieved as change with distance z.(I-SOP, A-SOP) (v, z) value is now different along with the variation of z in whole OTDR, rather than a number difference only, and provides DGD to the different Δ T that obtains with different _RoundTrip(v, z).Notice that (v z) is the standardization track that changes along with fiber lengths z to T.

Yet, in this area, use very short pulse normally unpractical, because the Measuring Time that requirement is extremely grown between the useful dynamic area in ground to ground.And, owing to the high-caliber coherent noise that uses short pulse to cause reduce may require one big to unacceptable equivalent laser linewidth, it causes little maximum detection amount PMD amount.The present invention considers following discovery: with big long pulse, mean square deviation＜Δ T (v, z) ² _SOP" dwindled " factor simply, this factor can calculate from be independent of identical raw data.(notice that, subscript SOP is illustrated in that (I-SOP is A-SOP) to last average here.) factor sigma may _r ²(z v) is the relative variance of these tracks, and it is the function of z---depend on the local characteristic of this optical fiber, be defined as,

σ_{r}^{2} (z, v) = {(\frac{1}{σ_{10}})}^{2} [{< T {(z, v)}^{2} >}_{SOP} - {< T (z, v) >}_{SOP}^{2}] - - - (14)

Wherein, with reference to variance be

σ_{10}^{2} = 4 / 45 .

Then, by with the mean square deviation in the equation (12) divided by the relative variance in the equation (14), obtain at the round DGD that gives fix midway point wavelength place, promptly

{DGD}_{RoundTrip} (z, v) = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{\frac{{< ΔT {(z, v)}^{2} >}_{SOP}}{σ_{r}^{2} (z, v)}}) - - - (15)

In addition, in a preferred embodiment of the invention, preferably, at (I-SOP, A-SOP) to centre wavelength pairing two tight alternate set of wavelengths should constantly change, then equation (14) and (15) are carried out indicated on average, obtain thus to come and go PMD, rather than the specific DGD that the certain wave strong point only obtains.

{PMD}_{RoundTrip} (z) = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{\frac{{< ΔT {(z, v)}^{2} >}_{SOP; λ}}{σ_{r}^{2} (z)}}) - - - (16)

In addition, because the general user is more prone to PMD value forward rather than round value are shown, so this result is multiplied by the above-mentioned round factor

α_{rt} = \sqrt{3 / 8} .

Like this, PMD is forward,

PMD(z)＝α _rt·PMD _RoundTrip(Z) (17)

Wherein in the equation (14) in that (I-SOP A-SOP) on average also is replaced with in that (I-SOP is A-SOP) to asking average with wavelength on the two, promptly last asking

σ_{r}^{2} (z) = {(\frac{1}{σ_{10}})}^{2} [{< T {(z, v)}^{2} >}_{SOP; λ} - {< T (z, v) >}_{SOP; λ}^{2}] - - - (18)

It should be noted that by the round DGD that gives fix midway point wavelength place is asked root mean square or average on the range of wavelengths of regulation, also can obtain to come and go rms DGD or come and go average DGD (mean DGD _RoundTrip) (promptly coming and going PMD), for

rms {DGD}_{RoundTrip} (z) = \sqrt{{< {DGD}_{RoundTrip} {(z)}^{2} >}_{λ}}

And

mean?DGD _RoundTrip(z)＝<DGD _RoundTrip(z)> _λ

Then, by respectively to rms DGD _RoundTripWith average DGD _RoundTripMultiply by a round factor simply

With 2/ π, can obtain rms DGD (z) and average DGD (z) forward.

Under the restriction of the enough little optical frequency poor (" frequency step ") between the closely alternate wavelength, equation (15) and (16) are reduced to following differential formulas simply,

{DGD}_{RoundTrip} (z, v) = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< ΔT {(z, v)}^{2} >}_{SOP}}{σ_{r}^{2} (z, v)}} - - - (15 a)

{PMD}_{RoundTrip} (z) = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< ΔT {(z, v)}^{2} >}_{SOP; λ}}{σ_{r}^{2} (z)}} - - - (16 a)

Should note, another kind of possibility is (though be not very ideal alternative), it will also be appreciated that, in above-mentioned equation (8), (11), (13) and (16), at (I-SOP, A-SOP) to wavelength on ask only on average can be replaced with on big interval optical frequency (being wavelength) ask average, wherein (I-SOP is A-SOP) to keeping constant.Yet, under this " constant SOP " situation, this method has lost its applicability to all FUT types, promptly, if only the mid point wavelength is scanned, and (I-SOP is not A-SOP) to disturbing partially to used, these relations are just no longer effectively general so, even the fiduciary level of measurement and/or degree of accuracy are obviously lower---and still effective roughly.Usually, do not disturb partially if do not carry out, then this method have only when FUT be " ideal " or " near ideal ", and it is just effective when having selected big PMD Δ v value (being typically greater than 10)---wherein Δ v is the width in optical frequency interval---, described " ideal " or " near ideal " are optical fiber and show remarkable coupling at random, and have limited or " being close to limited " polarization coupled rate.Therefore, in fact, can not under a rational uncertainty, record little PMD value.In addition, people usually wish the early stage optical fiber of installing is measured, and the optical fiber since these optical fiber and about calendar year 2001 is compared usually more not " ideal ".

Should note, for this restriction of enough little optical frequency poor (" frequency step ") between the closely alternate wavelength, the equation of above-mentioned and following DGD that is used to calculate or PMD---as simple differential formulas---is basic equation, and, in order to obtain the optimal representation of this instrument, obtain big " frequency step " inverse sine formula from simple differential formulas.

It should be noted that also above-mentioned and following use relative variance calculates the equation of DGD or PMD, applicable to standardization power (comprising standardization OTDR track) and relative power (comprising relative OTDR track).It should be noted that also relative power (or relative OTDR track) is proportional with standardization power (or standardization OTDR track).

Method of operating, data processing and calculating

The common base that both-end PMD measures, single-ended overall PMD measures and single-ended accumulation PMD measures is " be used for the polarization state of input and output at random that PMD measures and disturb inclined to one side analysis (SSA) ", but they are different to the detail operations of those designed instruments.For example, double-end measurement must input optical controller device be placed on the end of FUT, and analyzer and pick-up unit is placed on the other end of FUT.Applied light source also possibility is different, for example, not only can use continuous wave light source but also can use light-pulse generator if both-end PMD measures---can select, or the optical frequency of light modulated, so that measure two or three tight alternate wavelength of generation for this; But measure for single-ended PMD, then be necessary, to solve reflection from the FUT far-end with light-pulse generator (normally tunable OTDR).Even the single-ended PMD for overall PMD and accumulation PMD measures, they are in pulse length, closely still have different slightly operations aspect quantity, the data of being obtained and the data processing of alternate wavelength.

Therefore, we will measure and single-ended accumulation PMD measurement description method of operating, data processing and calculating at both-end PMD measurement, single-ended overall PMD in three different chapters and sections below.

Method of operating

Method of operating at both-end PMD measurement

Now with reference to the process flow diagram shown in Fig. 4 A, 4B, 4C and the 4D, the method for operating on measurement DGD and/or PMD of both-end PMD surveying instrument shown in Figure 1 is described in more detail.In step 4.1 and 4.2, the user at first installs application software, and test module inserted platform, then start testing software so that this system initialization test module, specifically, be initialization wavelength polarized light source 12 (tunable laser source 12A or wideband light source 12B), input SOP controller 14A, analytical equipment 14B and 20 and detecting device 22 and processing element 34.Then, one end of tested optical fiber (FUT) 18 will be connected to the coherent source module, after input SOP controller 14A, and the far-end of FUT 18 will be connected to analyzer module, and the patch cord with PC or APC connector (such as FC/PC or FC/APC) is used to make these modules to be connected with this FUT.Most of instrument parameters are set up in factory according to customer requirement usually, but the user can pass through step 4.1c and 4.3 respectively, manually are that light source and analyzer are selected parameter.Suppose that the user has selected the manual parameters setting, this program advances to manual parameters step 4.1c and 4.4 is set so, and the prompting user is as follows:

(a) centre wavelength is set for tunable laser source 12A or tunable optic filter 27.

(b) for that group switching centre wavelength an interval [λ who covered in the light source 12 is set _Min, λ _Max], condition is the inserted range of wavelengths of this interval corresponding to FUT 18.

(c) be arranged to right tight alternate optical frequency v _UAnd v _LStride between (or wavelength) or difference δ v (or δ λ) be not if can (promptly fix in factory).Alternatively, the user can import the expection PMD value of this FUT, waits until processor and calculates, and selects wavelength (being optical frequency) stride then.For example, this stride can be set to δ v=α easily _{δ v}PMD ^-1, α wherein _{δ v}Approximate 0.15 to 0.2, so just can be from δ λ ≈ (c/v _c ²) δ v draws δ λ, wherein v _c=(v _U+ v _L)/2.(note: for given PMD value, have best stride, it should be big as far as possible so that signal to noise ratio (S/N ratio) maximizes, but enough little of to satisfy above-mentioned condition, promptly PMD δ v is less than 0.15 to 0.2.It should be noted that also closely alternate optical frequency (or wavelength) also can be more than two, this to the DGD of FUT wherein or PMD may time to time change test and monitoring especially meaningful.)

(d) centre wavelength is set and/or the quantity K of the polarization state selected by I-SOP scrambler 14A and A-SOP scrambler 14B, the quantity of data set promptly to be obtained (K).For example, K can be set to 1000 to 100,000.Or, alternatively,, only need to be provided with the number K of centre wavelength, and then be set sweep time for importing SOP controller and

analytical equipment

14B and 20 for continuous sweep input and output SOP pattern.Perhaps, alternatively,, only need be provided with by the quantity K of the polarization state of I-SOP scrambler 14A and A-SOP scrambler 14B selection if only selected a centre wavelength, or the sweep time of I-SOP scrambler 14A and the two continuous sweep of A-SOP scrambler 14B.

(f) alternatively, if be provided with a series of modulating light pulses in FUT, setting will be by the number of average lasting pulse, to obtain each individual power (for example 2 or greater than 100).If only have a modulating light pulse to be launched among this FUT, then do not require any setting.

(g) measure number of times for each individual PMD measurements is provided with overall total acquisition time and PMD, with and how two measurements in office between stand-by period.

(h) the duration T p of the light pulse of selection modulation.Typically, select long pulse length for measuring, because it causes between high dynamic area and high s/n ratio, although short pulse still can use.(typically, modulating light pulse is selected between the 100 μ s to 1s, although this interval pulse length in addition is also feasible.)

The power input of tunable optical source apparatus is set (i) alternatively.

(j) alternatively, by on this light path---for example just a position after the input end of analyzer module---optical attenuator is regulated the power that enters analyzer module from FUT.But this automatically is provided with by this instrument usually.

(k) alternatively, the title and/or the relevant information of input optical cable or optical fiber.

(l) all measurement parameters are saved in data file, this data file will be obtained to be used for data processing by data processor 34.

If at determination step 4.3, the user selects the Automatic parameter setting, the following steps of the Automatic parameter setting up procedure of this program start step 4.5, and execution so:

(a) select predetermined specific acquiescence measurement parameter, promptly

(1) the centre wavelength interval [λ that is covered by light source 12 _Min, λ _Max],

(2) obtain for PMD data, the SOP due to I-SOP scrambler 14A and the A-SOP scrambler 14B and/or the quantity K of centre wavelength, or alternatively, the sweep time of I-SOP scrambler 14A and A-SOP scrambler 14B,

(3) be used for time, any twice individuality that each individuality obtains (measurement) between obtaining stand-by period and the number of times that repeats to obtain,

(4) the frequency pulse duration T p (or length) of tunable coherent source, and

(5) luminous power of being sent and the power that is received.

(b) test module also can be designed as and has pre-scanning process---use less group number, such as K=50 to 100, to obtain to two tight alternate optical frequency v _UAnd v _L(or wavelength X _UAnd λ _L) between the estimation of optimal wavelength stride difference on the frequency δ v (or δ λ).From carrying out the data that prescan obtains, to find two tight alternate optical frequency v _UAnd v _L(or wavelength X _UAnd λ _L) between suitable stride or difference δ v (frequency) or δ λ (wavelength).For example, this data are obtained and can be carried out in the following manner: to each group, by using four different optical maser wavelengths, obtain altogether the combination of 6 different frequencies or wavelength stride.In this case, may require between these FUT two ends suitable communication is arranged.

(c) also automatic mode be can design, automatic real estate third contact of a total solar or lunar eclipse cable or optical fiber title and/or relevant information come;

In case measurement parameter is transfused to, no matter be manually or automatically, this program advances to step 4.6, and calculate wavelength stride δ λ (if or difference on the frequency δ v)---total PMD of the expection of this FUT is designated or estimated by above-mentioned automatic setting up procedure, and is provided with based on this parameter and calculates suitable wavelength order λ _sPreferably use three or four (even more) different optical maser wavelengths to produce three or six (even more) different wavelength strides, with the measured PMD interval of covering wide.

Finally, all these measurement parameters, though be direct appointment or calculate as mentioned above, be stored in the header of data file or in the instrument (step 4.7).

The live width that it should be noted that tunable coherent source is arranged on relatively little level (for example less than 1 to 2GHz) usually in factory or by design, to guarantee to measure the ability of the high PMD (for example greater than 50ps) from FUT.

It should be noted that conventionally, in each SOP and/or central wavelength, two tight alternate optical frequency v _UAnd v _L(wavelength X _UAnd λ _L) between difference on the frequency δ v (or wavelength stride δ λ) can keep identical or close.Each SOP and/or wavelength can only be set in the time period of a weak point.

Fig. 4 (C) illustrates in greater detail the data acquisition step 4.10 that is used to obtain k group power.Predetermined wavelength stride δ λ can be used to calculate wavelength order λ _s, as discussing in step 4.6.Frequency v _L ^(k)And v _U ^(k)Calculated, satisfied v _L ^(k)-v _U ^(k)=δ v, wherein δ v is that difference on the frequency (perhaps when wavelength difference δ λ is used, satisfies λ _U ^(k)-λ _L ^(k)=δ λ).Corresponding to the maximum detection amount PMD of given stride δ v, i.e. PMD _Max, can be estimated as PMD _Max～α _Rt(π δ v) ^-1, and δ v can from

δλ = (λ_{0}^{2} / c) \cdot δv

Draw, wherein λ ₀=(λ _Min+ λ _Max)/2.Control module 30 control test modules, as follows to obtain k group power:

● by I-SOP scrambler 14A and A-SOP scrambler 14B SOP is set _k(the step 4.3.1 of Fig. 4 (C)) if---macroscopical SOP stride selects to be applied to

scrambler

14A, 14B one or both of; If perhaps SOP scanning is applied to

scrambler

14A, 14B one or both of continuously, then be set sweep time for input and

output scrambler

14A, 14B, wherein input and output SOP can be by scanning randomly continuously lentamente, to cover Poincare sphere equably.

● control light source 12 or tunable optic filter 27 are set to λ to hang down wavelength _L ^(k)(the step 4.3.2 of Fig. 4 C).Detection and processing unit 34 will obtain power data such as P _XLAnd P _YL(the step 4.3.3 of Fig. 4 C).The more details of this data acquisition shown in Fig. 4 D will be described below.Identical data acquisition is repeated, to obtain second part or the power P that repeats _XL" and P _YL" (the step 4.3.4 of Fig. 4 C)

● at high wavelength X _U ^(k)(λ wherein _U ^(k)Also be provided with by light source 12 or tunable optic filter 27) repeat identical data acquisition, be that the two control of I-SOP scrambler 14A and A-SOP scrambler 14B keeps approximately uniform input and output SOP simultaneously.Then, detection and processing unit 36 obtain power data P _XUAnd P _YU, and the P that repeats _XU" and P _YU" (the step 4.3.5 of Fig. 4 C, 4.3.6 and 4.3.7), or alternatively, can obtain this data from the time period of a weak point, but these data are divided into two data that come across different time sections.

Fig. 4 D has provided being used for shown in Fig. 4 C and has obtained the P of k group power _XLAnd P _YLThe more details of data acquisition of step 4.3.3.The modulating light pulse that is sent from light source 12 is admitted to FUT 18, and the output modulating light pulse comes out from the far-end of FUT18.Then, the modulating light pulse that comes out is admitted to the test analyzer module of instrument, so that---one of two output arm is connected with linear polarization 20A---is divided into two-way, y and x by PBS 20 or 20C or coupling mechanism 21, for example three-dB coupler.What enter road y and road x cuts apart light pulse by two OPTICAL SENSORS---two APD for example, such as 22B and 22C (or 20)---detect (the step 4.4.1 of Fig. 4 D and 4.4.2).Alternatively, the described modulating light pulse that comes out that incides the test analyzer module is sent straight to linear polarization.This light pulse is directly by an OPTICAL SENSORS---APD for example, such as 22A---detect (Figure 1B),---for example three-dB coupler---is divided into two-way perhaps to be coupled device 21, y and x, this light pulse enters road y and road x, and---for example two APD such as 22B and 22C---detects (Fig. 1 H) by two OPTICAL SENSORS.Be sampled or sample and on average from " continuing " response signal of the modulating light pulse of FUT far-end, to obtain " response impulse signal ", such as P _y(t) and P _x(t) (the step 4.4.3 of Fig. 4 D and 4.4.4).Then,---in its duration ask average on most time of the pulse center of impulse response signal---obtains also average power P of final sampling or sampling by the described response impulse signal that had before obtained is averaged _XLAnd P _YL(the step 4.4.5 of Fig. 4 D and 4.4.6).The length for the treatment of the average duration of pulse depends on the pre-filtering of electron device usually.

In case k group power (is seen Fig. 4 B) and is acquired as described above that then in step 4.11, the data of k group are saved in the data file in step 4.10.Then, step 4.12 makes the increment of group number register.

Step 4.4 is set or step 4.5 or default parameters setting is set according to manual parameters by Automatic parameter, to different centre wavelength and/or I-SOP scrambler 14A and the selected input and output SOP of A-SOP scrambler 14B, repeating data obtaining step 4.10 and group storing step 4.11 all are acquired and are stored in the data file up to K power packages.

Whether step 4.9 will determine current individuality to obtain to finish.If determination step 4.9 has provided positive result, then this program is in step 4.11 storage data.If do not finish, then this acquisition process is incited somebody to action execution in step 4.10 and 4.11 once more.

Step 4.8 will determine whether to start once new individual bulk measurement to be obtained.If whole measurement is obtained and finished, then step 4.15 will be preserved all individual data items, be used for overall complete obtaining.If do not finish, then processor obtains replacement k=0 to start once new individuality, is used for step 4.9,4.10,4.11 and 4.12.Step 4.16 will determine whether to start another time to be obtained.

In this stage, measurement parameter and all power packages have been stored in the suitable file.

Determination step 4.17 can start data processor; Step 4.18 can load the current available data of obtaining from data file; Step 4.19 can be handled these data, estimating this FUT in the DGD of given central wavelength value, or average DGD on a range of wavelengths or rms DGD; Step 4.21 can show it.Alternatively, step 20 can allow the user to preserve handled result, such as DGD or average DGD or rms DGD value and time relation.

Then, can give chance of user from the optional judgement of step 4.16 beginning initiates processing is obtained in another time of same FUT.If the user determines to do like this, this program turns back to parameter step 4.3 is set so.Do not do like this if do not determine, determination step 4.17 allows the user to select to withdraw from acquisition process, in this case, be stored in data in the data file and will be held being used for processing after a while, or determination step 4.17 allows users to initiate being acquired and the processing of stored power data.

Initiated if handle, then step 4.18 allows the user to select pending data file in the Open File dialog box of routine, data processor 34 by the visit of this data file previous preserve obtain data (it comprises power and the related measurement parameter that is detected), and use these data to calculate DGD or average DGD or the rms DGD of this FUT.

Should note, above-mentioned steps can obtain rms DGD (being PMD) and obtain at the DGD at given endpoint wavelength place or the DGD that changes along with wavelength, and can calculate rms DGD or equal DGD with the method that following chapters and sections will be described, this method also can be included in the data processing step 4.19.

Note, situation for K=1, promptly, for group only with identical input and output SOP and identical centre wavelength, the power of light can be obtained in a similar fashion, also can estimate PMD roughly for this situation, although this simple situation may not provide enough accurate and significative results, because there is significant uncertainty in the possibility of result that records.

Method of operating at single-ended overall PMD measurement

Now with reference to the process flow diagram shown in Fig. 5 A, 5B and the 5C, the method for operating that the single-ended PMD based on tunable OTDR shown in Fig. 2 G and the 2C measures is described.In step 5.1, the user at first installs application software, and test module is inserted platform, starts testing software then so that this system initialization test module, specifically, be that the tunable light-pulse generator 12 of initialization, I/O-SOP controller 14 and OTDR detect and processing element 34.Then, tested optical fiber (FUT) 18 will be connected to test module (being instrument), and have the patch cord of PC connector (such as FC/PC or FC/UPC) or the far-end that optical fiber pigtail mirror 50 is connected to this FUT.This will cause the local reflection in the end of FUT, and this reflection is used to carry out PMD and measures.

Determination step 5.2 prompting users select manual parameters setting or Automatic parameter setting.Suppose that the user selects the manual parameters setting, then this program advances to manual parameters step 5.3 is set, and the prompting user is as follows:

(a) the group switching centre wavelength is provided with the range of wavelengths [λ that will be covered by tunable pulsed laser source 12 _Min, λ _Max].

(b) two tight alternate optical frequency v are set _UAnd v _LFrequency step between (or wavelength) or difference on the frequency δ v (or δ λ).Alternatively, the user can import the expection PMD value at this FUT, and waits until processor 34 and select the wavelength stride.For example, this stride can be set to δ v=α easily _{δ v}PMD ^-1, α wherein _{δ v}Approximate 0.1 to 0.15, like this, δ λ can be from δ λ ≈ (c/v _c ²) δ v draws, v wherein _c=(v _U+ v _L)/2.(note: have best stride for a given PMD value, its big as far as possible so that signal to noise ratio (S/N ratio) maximizes, but enough little be that PMD δ v is less than 0.1 to 0.15 to satisfy above-mentioned condition.)

(c) quantity K centre wavelength and/or that pass through the polarization state of I/O-SOP scrambler 14 selections, the quantity of data set promptly to be obtained (K) are set.For example, K can be set to 200.

(d) Δ t averaging time (for example, Δ t equals 0.05 or 0.10 second) of each individual power is set, or lasting number of pulses from this FUT far-end reflection (for example 50 or 100) is set.Note, be provided with centre wavelength averaging time Δ t and quantity K and/or polarization state after, also can obtain total acquisition time that PMD measures.

(e) duration of pulse Tp (such as equaling for 275,1000,2500,5000,10000,20000 nanoseconds) or the pulse length of selection OTDR.Do not overlap this pulse length L from some of the pulse of another reflection in order to make reflection in time from the pulse of selected reflection with reflection _pShould be selected as L _p＜Δ z, wherein Δ z is along this FUT, in selected reflection and nearest any reflection or whole distances between other reflections.Typically, be that single-ended PMD measures to select long pulse length because it have make between dynamic area high and/or make signal to noise ratio (S/N ratio) high and/or make the advantage of the time of asking average short (make thus overall acquisition time short), although short pulse still can be used.

(f) FUT length is set, normally the optical length in full force and effect of this FUT.

(g) alternatively, select between high dynamic area according to fiber lengths or between low dynamic area.Typically, in common operation, test module prompting user selects between high dynamic area, but also can allow between low dynamic area that user by selecting is used to obtain with the very short optical fiber of test.Use low dynamic area inter mode, the output peak power of the OTDR pulse of being started is reduced, perhaps by in light path, inserting optical attenuator, for example, the position before the output terminal of test module just, perhaps electrically, for example, the bias current of the gain media by reducing this tunable pulsed laser device.

(h) alternatively, the title and/or the relevant information of input optical cable or optical fiber.

(i) all measurement parameters are saved in data file, this data file will be obtained to be used for data processing by data processor 34.

If at determination step 5.2, the user selects the Automatic parameter setting, then this program starts the Automatic parameter setting up procedure in step 5.4, and carries out following steps:

(a) select predetermined particular default measurement parameter, promptly

(6) the centre wavelength interval [λ that will be covered by tunable pulsed laser source 12 _Min, λ _Max],

(7) remain by I/O-SOP controller 14 be provided be used for actual single-ended PMD data acquisition (I-SOP, A-SOP) to and/or the quantity K of centre wavelength (for example 200),

(8) each individual power ask average time Δ t (for example, Δ t equals 0.05 or 0.1 second) or from the quantity of the lasting pulse of this FUT far-end reflection (for example 50 or 100), and

(9) the duration of pulse T of OTDR _p(or length).

Notice that these default parameterss that are provided with also will be used to prescan and obtain in (1), (3) and (4).

(b) test module group number that use is reduced such as K=50, carries out prescan and obtains, with estimate FUT length, from total losses and two tight alternate optical frequency v of FUT _UAnd v _L(or wavelength X _UAnd λ _L) between optimal wavelength stride difference on the frequency δ v (or δ λ).OTDR is emission standard OTDR pulse (for example 1 or 10 microseconds), with the end of detecting this optical fiber (or user limit local reflection), so that can obtain FUT length, and can also be by through deriving pulse repetition time (T the two-way time of this fiber lengths _r).Obtain from this OTDR, also can estimate the loss of FUT, otherwise can observe the state of saturation on the OPTICAL SENSORS---if present.Then, can automatically determine whether to reduce the output peak power of this OTDR light source.Prescanned data obtains and is performed, and can find two tight alternate optical frequency v _UAnd v _L(or wavelength X _UAnd λ _L) between suitable frequency step δ v (difference on the frequency) or δ λ (wavelength difference).For example, this data are obtained and can be carried out in the following manner: to each group, by using four different optical maser wavelengths, obtain altogether the combination of 6 different frequencies or wavelength stride.Obtain power data by handling these prescans, can find the best appropriate wavelength stride of waiting during being used in actual single-ended PMD measurement data obtains.All automatically selected measurement parameters are saved in the header of data file, and this data file will be obtained for data processor 34 carries out data processing.

In case measurement parameter is transfused to, no matter be manually or automatically, this program advances to step 5.5, and calculate wavelength stride δ λ (if or difference on the frequency δ v)---total PMD of the expection of this FUT is designated or estimated by above-mentioned automatic setting up procedure; And according to calculating repetition period Tr through two-way time of this fiber lengths; And be provided with based on this parameter and calculate suitable wavelength order λ _s

Finally, all these measurement parameters, though be direct appointment or calculate as mentioned above, be stored in the header of data file (step 5.6).

The live width that it should be noted that tunable light-pulse generator is provided with in factory usually, is arranged on relatively little level (for example 1-2GHz or littler), to guarantee to measure the ability from the high PMD of FUT.

When the group number register was initialized to k=0, determination step 5.7 determined whether total power packages number is acquired.Were it not for, then this program advances to step 5.8, to obtain k group power.

It should be noted that conventionally, in each SOP and/or central wavelength, two tight alternate optical frequency v _UAnd v _L(wavelength X _UAnd λ _L) between difference on the frequency δ v (or wavelength stride δ λ) can keep identical or close.Each SOP and/or wavelength can only be set in the short time period.

Fig. 5 B illustrates in greater detail the data acquisition step 5.8 that is used to obtain k group power.Predetermined wavelength stride δ λ can be used to calculate wavelength order λ _s, as discussing in step 4.5.Frequency v _L ^(k)And v _U ^(k)Calculated, satisfied v _L ^(k)-v _U ^(k)=δ v, wherein δ v is that difference on the frequency (perhaps when wavelength difference δ λ is used, satisfies λ _U ^(k)-λ _L ^(k)=δ λ).Corresponding to the maximum detection amount PMD of given stride δ v, i.e. PMD _Max, can be estimated as PMD _Max～α _Rt(π δ v) ^-1, and δ v can from

δλ = (λ_{0}^{2} / c) \cdot δv

● by the I-SOP controller SOP is set _k(the step 5.3.1 of Fig. 5 B).

● control tunable pulsed laser 12 is set to λ to hang down wavelength _L ^(k)(the step 5.3.2 of Fig. 5 B).Detection and processing unit 36 will obtain power data such as P _XLAnd P _YL(the step 5.3.3 of Fig. 5 B).The more details of this data acquisition shown in Fig. 4 C will be described below.Identical data acquisition will be repeated, to obtain second part or the power P that repeats _XL" and P _YL" (the step 5.3.4 of Fig. 5 B)

● at high wavelength X _U ^(k)(λ wherein _U ^(k)Also be provided with by tunable laser 12) repeat identical data acquisition, keep simultaneously identical (I-SOP, A-SOP) right.Then, detection and processing unit 36 obtain power data P _XUAnd P _YU, and the P that repeats _XU" and P _YU" (the step 5.3.5 of Fig. 5 B, 5.3.6 and 5.3.7).

Fig. 5 C has provided being used for shown in Fig. 5 B and has obtained the P of k group power _XLAnd P _YThe more details of data acquisition of step 5.3.3.The modulating light pulse that is sent from OTDR is admitted to FUT 18, and sub-fraction (or most) pulsed light reflects from the local reverberator---for example, and by PC connector that uses patch cord or the optical fiber pigtail mirror that is connected FUT end place.Then, reflection light pulse is returned in test module or the instrument, so that---one of two output arm is connected with linear polarization---is divided into two-way, y and x by PBS or coupling mechanism, for example three-dB coupler.What enter road y and road x cuts apart light pulse by two OPTICAL SENSORS---and two APD for example, detect (the step 5.4.1 of Fig. 5 C and 5.4.2) such as 22 ' B and 22 ' C.From the FUT far-end or be sampled and on average from " continuing " response signal of the reflection light pulse of any other position on the optical fiber, to obtain the average response pulse signal of " through average ", such as P _y(t) and P _x(t) (the step 5.4.3 of Fig. 5 C and 5.3.4).Then,---in its duration ask average on most time of the pulse center of impulse response signal---obtains final average power P by averaging to described previous sampling and through average response impulse signal _XLAnd P _YL(the step 5.4.5 of Fig. 5 C and 5.4.6).Treat that average pulse persistance length depends on the pre-filtering of electron device usually.

In case k group power (is seen Fig. 5 A) and is acquired as described above that then the data of k group are saved in the data file in step 5.9.Then, step 5.10 makes the increment of group number register.

Step 5.3 is set or step 5.4 is set according to manual parameters by Automatic parameter, (I-SOP to different centre wavelength and/or the selection of I/O-SOP controller, A-SOP) right, repeating data obtaining step 5.8 and group storing step 5.9 all are acquired and are stored in the data file up to K power packages.

In this stage, measurement parameter all has been stored in the identical data file with all power packages, and this document is related with the header information of measurement parameter.

In data acquisition step 5.20 (optional), can load the current available data of having obtained from data file, and handle them, to estimate rms DGD (the being PMD) value of FUT 18, step 5.21 can show that it and this obtain the time of exhausting, length and the loss of this FUT.Notice that estimated PMD value can frequently change, and obtains end up to data.Alternatively, step 5.22 can allow the user to preserve result.

Also be in this stage, determination step 5.7 has provided positive result, and in step 5.11, this program Saves and Closes data file.

Then, can give chance of user from the optional judgement of step 5.12 beginning and initiate the chance of obtaining another power packages K of same FUT.If the user determines to initiate, then this program turns back to parameter step 5.2 is set.Do not do like this if do not determine, determination step 5.13 allows the user to select to withdraw from acquisition process, in this case, be stored in data in the data file and will be held being used for processing after a while, or determination step 5.13 allows users to initiate being acquired and the processing of stored power data.

Initiated if handle, step 5.14 allows the user to select pending data file in " opening file " frame of routine, so, in step 5.16, the measurement parameter that obtains power data and association that data processor 34 had before been preserved from this data file visit, and use these data to calculate total rms DGD (being PMD) of this FUT.On the other hand, indication, the user can start data processing software to square frame 5.15---it is not similar " step "---at any time independently, to handle any data file of before having obtained.In step 5.17, data processor 34 will calculate PMD value result and measurement parameter preserve hereof, in step 5.18, show or PMD value that output records that length and the loss of other results such as this FUT can also be arranged.

Note, situation for K=1, promptly, for only one have identical (I-SOP, A-SOP) to and the group of identical centre wavelength, the light reflective power can be obtained in a similar fashion, also can estimate PMD roughly, although this simple situation may not provide enough accurate and significative results, because there is significant uncertainty in the possibility of result that records.

Data processing step 5.16 is handled the mode of the data of being stored chapters and sections is below described.

Should note, above-mentioned steps can obtain rms DGD (being PMD), but also the DGD that can obtain to change along with wavelength can calculate rms DGD or average DGD with the method that following chapters and sections are described then, and this method also can be included in the data processing step 5.16.

Method of operating at single-ended accumulation PMD measurement

Now with reference to the process flow diagram shown in Fig. 6 A and the 6B, the POTDR method of operating that the accumulation PMD that the length with FUT changes is measured in shown in Figure 3 being used to is described.In step 6.1, the user makes this system initialization POTDR, specifically, is that the tunable light-pulse generator 12 of initialization, I/O-SOP controller 14 and OTDR detect and processing element.Determination step 6.2 prompting users select manual parameters setting or Automatic parameter setting.Suppose that the user has selected the manual parameters setting, then this program advances to manual parameters step 6.3 is set, and the prompting user is as follows:

(b) be arranged to right tight alternate optical frequency v _UAnd v _LFrequency step between (or wavelength) or difference δ v (or wavelength difference δ λ).Alternatively, the user can import the expection PMD value of this FUT, and waits until processor 34 and select the wavelength stride.For example, this stride can be set to δ v=α easily _{δ v}PMD ^-1, α wherein _{δ v}Approximate 0.1 to 0.15.It should be noted that this POTDR can be configured to allow the user to select stride number M greater than 1; Then, this control program will be selected M stride based on total PMD of the expection of this FUT, there is suitable ratio (to note: have best stride between these strides for given PMD value, it is big as far as possible so that the signal to noise ratio (S/N ratio) maximization, but enough little of to satisfy above-mentioned condition, promptly PMD δ v is less than 0.1 to 0.15.But facility described here must be carried out challenging task, promptly measures the big interval accumulation PMD value that changes along with z, the PMD=0 during from z=0, the total PMD of PMD=of the FUT during to z=FUT length simultaneously.Why Here it is needs a lot of measurements under different strides, so that measure FUT all different " sections " to obtain similar relative measurement degree of accuracy (for example recently representing) with percentage, perhaps as here or mentioned above, every group is used the tight alternate wavelength more than two, every group of N _λIndividual wavelength has caused that M=N is arranged in each scanning in theory _λ(N _λ-1) pair different strides/2, thus save time.)

(c) quantity K centre wavelength and/or that pass through the polarization state of I/O-SOP controller 14 selections, the quantity of trajectory set promptly to be obtained (K) are set.

(d) the average time Δ t of asking of each individual track (for example, Δ t equals 1 or 2 second) is set, or the quantity treat average electrical impulse response signal (for example 1250 or 2500) is set, to obtain each individual track.

(e) be set (as T the duration of pulse _pOr length=50,100,200,300 nanoseconds).

(f) specify FUT length, the normally optical length in full force and effect of this FUT.

If in step 6.2, the user has selected the Automatic parameter setting, then this program advances to step 6.4, and carries out following steps:

● select specific acquiescence measurement parameter, promptly

(1) the centre wavelength interval [λ that will be covered by tunable pulsed laser source 12 _Min, λ _Max], the whole range of wavelengths that actual typically tunable laser can obtain,

(2) by I/O-SOP controller 14 being used for of being provided with centre wavelength that final POTDR data obtain and/or (I-SOP, A-SOP) right quantity K, for example 100 or 200,

(3) each individual POTDR track ask Δ t averaging time (for example, Δ t equals 1 or 2 second), or treat the quantity (for example 1250 or 2500) of average electrical impulse response signal,

(4) duration of pulse (T for example _p=50,100,200,300 nanoseconds) or pulse length, and

(5) live width of tunable pulsed laser (optional).

Notice that these default parameterss that are provided with also will be used to prescan and obtain in (1), (3), (4) and (5).

This POTDR uses the group number---such as K=20---that reduces to carry out prescan, to estimate FUT length, two tight alternate optical frequency v roughly _UAnd v _L(or λ _UAnd λ _L) between the wavelength stride δ λ of optimization (or difference on the frequency δ v).Like this, this OTDR with starting standard OTDR pulse (for example 1 μ s) with the detection fiber end, so that can be according to through two-way time of fiber lengths obtaining FUT length and derive the pulse repetition time.Then, obtaining of OTDR track will be performed, and with via the Fast estimation to the overall PMD of this FUT, find two tight alternate optical frequency v _UAnd v _L(or λ _UAnd λ _L) between suitable stride of the best or difference δ v (or δ λ).For example, this data are obtained and can be carried out in the following manner: to each group, by using four different optical maser wavelengths, obtain the wavelength stride combination that 6 (M=6) is different altogether.By handling these prescanned datas, can find the best appropriate wavelength stride of waiting during being used in actual POTDR data obtains.

In case measurement parameter is transfused to, no matter be manually or automatically, this program advances to step 6.5, and calculate wavelength stride δ λ (if or difference on the frequency δ v)---total PMD of the expection of this FUT is designated or estimated by above-mentioned automatic setting up procedure; And according to calculating repetition period T through two-way time of this fiber lengths _rAnd be provided with based on this parameter and calculate suitable wavelength order λ _s

Finally, all these measurement parameters, though be direct appointment or calculate as mentioned above, all be stored in the header of data file (step 6.6).

If showing one, Fig. 6 A optionally is used for laser linewidth being set the step (following step 6.5) of---words that lasing light emitter 12 allows---according to the parameter of before importing.For example, little (big) live width can be selected in order to measure big (little) total PMD.Do not specifying total PMD and do not carrying out under the situation of automatic setting up procedure, specified wavelength stride (δ λ) can be used to estimate total PMD, also can select laser linewidth in view of the above then.

When the group number register was initialized to k=0, determination step 6.7 determined whether the sum of trajectory set is acquired; Were it not for, then this program advances to step 6.8, to obtain k group OTDR track.

Fig. 6 B illustrates in greater detail the track obtaining step 6.8 that is used to obtain k group OTDR track.As mentioned before, at two tight alternate optical frequency v _UAnd v _LHave at least one predetermined difference on the frequency δ v (or wavelength stride δ λ) between (or wavelength), therefore the sum of selected optical maser wavelength must be two at least.If use a plurality of different wavelength stride δ λ, then these wavelength strides can be selected, to measure the different interval of PMD value best.For example, can select to use two wavelength strides, δ λ ₁With δ λ ₂, this requires each group that N is arranged _λ=3 different wavelength.In addition, the wisdom of the ratio of described two strides select can be, for example, δ λ 1 _/δ λ ₂=5.Corresponding to the maximum detection amount PMD of given stride δ v, i.e. PMD _Max, can be estimated as PMD _Max～α _Rt(π δ v) ^-1, and δ λ can from

δλ = (λ_{0}^{2} / c) \cdot δv

Draw, wherein λ ₀=(λ _Min+ λ _Max)/2.Control module 30 control POTDR, as follows to obtain k group track:

● be provided with (I-SOP by I/O-SOP controller 14 _k, A-SOP _k) (the step 6.8.1 of Fig. 6 B).

● control tunable pulsed laser device 12, so that wavelength is set to λ _L ^(k)(the step 6.8.2 of Fig. 6 B) sends the OTDR light pulse then.Detection and processing unit 36 obtain OTDR track P _XLAnd P _YL(the step 6.8.3 of Fig. 6 B).Identical obtaining step is repeated, to obtain second part or the track P that repeats _XL" and P _YL" (the step 6.8.4 of Fig. 6 B).

● at high wavelength X _U ^(k)Repeat identical data acquisition, keep simultaneously identical (I-SOP, A-SOP).Then, detection and processing unit 36 obtain OTDR track P _XUAnd P _YUAnd the P that repeats _XU" and P _YU" (the step 6.8.9 of Fig. 6 B and 6.8.10).

● under this group comprises situation more than a pair of light pulse series, wavelength is set at least one additional wavelength X between low wavelength and the high wavelength _I ^(k)Under (the step 6.8.5 of Fig. 6 B).Detection and processing unit 36 obtain OTDR track P _XIAnd P _YI(the step 6.8.6 of Fig. 6 B).Identical data acquisition will be repeated, to obtain to repeat track P _XI" and P _YI" (the step 6.8.7 of Fig. 6 B).

In case k group OTDR track is obtained as described above, then (sees Fig. 6 A) in step 6.9, this group is saved in the data file.Then, step 6.10 makes the increment of group number register.

According to parameter step 6.2 or 6.3 are set, to different centre wavelength and/or the (I-SOP that selects by I/O-SOP controller 14 _k, A-SOP _k) right, repeating data obtaining step 6.8 and group storing step 6.9 all are acquired and are stored in the data file up to K group track.

In this stage, measurement parameter all has been stored in the identical data file with all OTDR trajectory set.

Also be in this stage, determination step 6.7 has provided positive result, and in step 6.11, this data file of this stop.Then, optionally determination step 6.12 gives chance of user and initiates the chance of obtaining to another trajectory set K of same FUT.If the user determines to initiate new acquisition process, then this program turns back to parameter step 6.2 is set.Do not do like this if do not determine, determination step 6.13 allows the user to select to withdraw from acquisition process, in this case, be stored in data in the data file and will be held being used for processing after a while, or determination step 6.13 allows users to initiate being acquired and the processing of stored power data.

If data processing is initiated, step 6.14 allows the user to select pending data file in the Open File dialog box of routine, so, in step 6.16, the measurement parameter that obtains data and association that data processor 34 had before been preserved from this data file visit, and use these data to calculate the accumulative total PMD that becomes with the distance on the FUT (z).On the other hand, indication, user's (even also not finishing any obtaining) at any time starts data processing software to square frame 6.15---it is not similar " step "---independently, to handle any data file of before having obtained.In step 6.17, data processor 32 is kept at result's (accumulation PMD curve and measurement parameter of for example changing along with z) in the obtainable file of spreadsheet, and in step 6.18, show or export the accumulation PMD curve of gained with tangible form.

Data processing step 6.16 is handled the mode of the data of being stored chapters and sections is below described.

Should note, above-mentioned steps can obtain rms DGD (being PMD), but also can obtain the DGD that changes along with wavelength, can calculate rms DGD or average DGD with the method for describing in the following chapters and sections then, this method also can be included in the data processing step 6.16.

Data processing and calculating

Data processing and calculating at both-end PMD measurement

Now data of description treatment step 6.19 is handled the mode of the data of being stored.

1. data structure

Constituted a primitive, that is, data are made of a performance number each luminous power from FUT---it is to obtain under given setting of described wavelength of method of operating that both-end PMD measures and input and output SOP---.Next data unit is one four power packages (i.e. four data unit), embodiment for Fig. 1 C and Fig. 1 G, be two covers, four power packages, wherein two power obtain from

OPTICAL SENSORS

22B and 22C simultaneously, and all obtain under the given input and output SOP of I-SOP scrambler 14A and A-SOP scrambler 14B setting.These two covers, four power packages that form the k group preferably obtain (time flows to right from a left side) with following sequence, or other similar fashion obtain---such as in the identical time but with the power (such as measuring identical power simultaneously with a coupling mechanism) of two repetitions of different detectors measure by two detecting devices, as:

I-SOP _k ^I, A-SOP _k ^OAnd/or λ _k:

The power that marks wherein that x and y refer to simultaneously or obtain from

OPTICAL SENSORS

22B and 22C respectively in the different slightly time, λ _L ^(k)-λ _L ^(k)Equal wavelength stride δ λ, the mid point wavelength is defined as

λ_{k} = (λ_{U}^{(k)} + λ_{L}^{(k)}) / 2,

The indication of two left-falling stroke symbol repeats power.

Finally, after obtaining, be stored in conceptual data in the data file and be drawn into matrix in the following equation (18), subsequently will be with reference to this matrix.This matrix comprises K group, and each group has four luminous powers (when using two OPTICAL SENSORS, two covers, four luminous power groups being arranged):

Should note, input SOP and output SOP can one by one selected at random (" macroscopical SOP strides "), or experience SOP scanning continuously slowly, under these two kinds of situations, all in the following way: along with the variation of time, each all covers Poincare sphere substantially equably.

2. the automatic calibration of relative gain

For the embodiment of Fig. 1 D, before proceeding any further calculating, be necessary to carry out the following calibration process of the relative gain of two detecting device 22B and 22C.For other embodiments,, then do not carry out this process if a detecting device is only arranged.

This calibrating principle is based on the following fact: the SOP that is used to generate enough big quantity when input and output SOP scrambler will come out with the ratio of 1: 1 (equating) two ports from PBS from the average power of the light of FUT 18 when covering Poincare sphere basically.Therefore, for viewed detecting device power, any viewed and departing from of this 1: 1 ratio can be quantized and include in consideration, and be as follows.

Data obtain finish after, K four luminous power groups that obtain from two OPTICAL SENSORS are stored, and, come J=4K altogether the power (data) of self-detector 22B that is, what come self-detector 22C also is J=4K power, describes as matrix (18).Below j the power (j=0,1... (J-1)) of 22B and 22C, be called as Px respectively _jAnd Py _jIf the overall loss in two arms of PBS equates, and the gain of two OPTICAL SENSORS and related electron device also equates, so power P _yAnd P _xRatio after asking on average on all J time will be

\frac{< Px >}{< Py >} &equiv; \frac{\underset{j}{Σ} {Px}_{j}}{\underset{j}{Σ} {Py}_{j}} = 1

But in practice, from the average of the power that records and ratio that obtain is not equal to 1, because the loss difference in two arms of PBS, and " effectively " of two OPTICAL SENSORS gain difference; Described " effectively " gain comprises the overall gain of photodiode response rate and following electron device---amplifier and sample circuit---.(note, needn't determine these individual gains respectively.) therefore, before proceeding all the other calculating, from all J power of OPTICAL SENSORS 22C acquisition, i.e. all Py _j, it is as follows all to make multiplication:

Py _j≡g _Forward·Py _j

Wherein

g_{Forward} = \frac{< Px >}{< Py >} = \frac{\underset{j}{Σ} {Px}_{j}}{\underset{j}{Σ} {Py}_{j}}

This calibration may be carried out at each wavelength place, but in practice, centre wavelength for those relatively closely alternate (for example less than 20nm), the relative wavelength dependency of member, detecting device or the like may be insignificant, and this calibration process is measured sequence to each PMD only to be needed to carry out once.

The result of this calibration is, promptly at all P _yPower (data) all multiply by after the relative gain that records as described above, and data processor 34 can the normalized luminous power.More accurately, under the situation of the embodiment of Fig. 1 D, standardization power is obtained as follows: with sample and the average signal P that comes self-detector 22B _x, or come the signal P of self-detector 22C _y, or (and being preferably) multiply by poor (Px-Py)/2 behind the weight coefficient or (Py-Px)/2 (will describe in detail at next chapters and sections), or any poor (1+w) that multiply by behind the weight coefficient ^-1(Px-wPy) (wherein w is a weighting factor), divided by come self-

detector

22B and 22C the two sample and average signal sum (Px+Py), should and expression be mapped to general power on this PBS, promptly do not select the general power of specific polarization component.

It should be noted that other calibrations also are feasible.For example, a kind of potential substituting collimation technique is: use the internal reference of fiber coupler (beam splitter) or inner reflector, sending to two different detecting devices from the luminous power of the predetermined amount (number percent) of the OTDR light of being launched.

To describe computing method below in detail to the standardization power of all preferred embodiments.

3. calculate

To the PMD value of Power Processing to obtain to describe now.It should be noted that hereinafter those symbols point to the matrix " data " in the equation (17) at all.Mark x and y refer to respectively from the back reflective luminous power of

OPTICAL SENSORS

22B and 22C acquisition.

3.1 standardization power

After this standardization power be noted as T, and it calculates according to the difference of each embodiment and difference.

(i) for the embodiment (two OPTICAL SENSORS connect same PBS) of Fig. 1 D, transmitance (standardization power) arbitrary calculating of following dual mode

T_{L}^{(k)} = \frac{{Px}_{L}^{(k)}}{{Px}_{L}^{(k)} + {Py}_{L}^{(k)}}

{T^{''}}_{L}^{(k)} = \frac{{Px}^{''}_{L}^{(k)}}{{Px}^{''}_{L}^{(k)} + {Py}^{''}_{L}^{(k)}}

T_{U}^{(k)} = \frac{{Px}_{U}^{(k)}}{{Px}_{U}^{(k)} + {Py}_{U}^{(k)}}

{T^{''}}_{U}^{(k)} = \frac{{Px}^{''}_{U}^{(k)}}{{Px}^{''}_{U}^{(k)} + {Py}^{''}_{U}^{(k)}} - - - (18 a)

Or

T_{L}^{(k)} = \frac{1}{2} \cdot \frac{{Px}_{L}^{(k)} - {Py}_{L}^{(k)}}{{Px}_{L}^{(k)} + {Py}_{L}^{(k)}}

{T^{''}}_{L}^{(k)} = \frac{1}{2} \cdot \frac{{Px}^{''}_{L}^{(k)} - {Py}^{''}_{L}^{(k)}}{{Px}^{''}_{L}^{(k)} + {Py}^{''}_{L}^{(k)}}

T_{U}^{(k)} = \frac{1}{2} \cdot \frac{{Px}_{U}^{(k)} - {Py}_{U}^{(k)}}{{Px}_{U}^{(k)} + {Py}_{U}^{(k)}}

{T^{''}}_{U}^{(k)} = \frac{1}{2} \cdot \frac{{Px}^{''}_{U}^{(k)} - {Py}^{''}_{U}^{(k)}}{{Px}^{''}_{U}^{(k)} + {Py}^{''}_{U}^{(k)}} - - - (18 b)

Wherein it will be appreciated that and be used in equation (18a) and (18b) before, different power P _yMultiply by the relative gain g that records in advance _Forward, as what in the description of automatic calibration process, indicate.

(ii) for the embodiment (two OPTICAL SENSORS connect same coupling mechanism) of Fig. 1 C, track P _xTo track P _yRatio at first be calculated as,

R_{L}^{(k)} = \frac{{Px}_{L}^{(k)}}{{Py}_{L}^{(k)}}

{R^{''}}_{L}^{(k)} = \frac{{Px}^{''}_{L}^{(k)}}{{Py}^{''}_{L}^{(k)}}

R_{U}^{(k)} = \frac{{Px}_{U}^{(k)}}{{Py}_{U}^{(k)}}

{R^{''}}_{U}^{(k)} = \frac{{Px}^{''}_{U}^{(k)}}{{Py}^{''}_{U}^{(k)}} - - - (18 c)

Then, above-mentioned ratio is by standardization---with respect to its on average and in addition standardization on K group, as follows:

T_{L}^{(k)} = u_{o} \frac{R_{L}^{(k)}}{{< R_{L} >}_{SOP}}

{T^{''}}_{L}^{(k)} = u_{o} \frac{{R^{''}}_{L}^{(k)}}{{< R_{L} >}_{SOP}}

T_{U}^{(k)} = u_{o} \frac{R_{U}^{(k)}}{{< R_{U} >}_{SOP}}

{T^{''}}_{U}^{(k)} = u_{o} \frac{{R^{''}}_{U}^{(k)}}{{< R_{U} >}_{SOP}} - - - (18 d)

Be u wherein with reference to average _o=1/2, thus average ratio R is defined as,

{< R_{L} >}_{SOP} = \frac{1}{2 K} \underset{k}{Σ} (R_{L}^{(k)} + R_{L}^{'' (k)})

{< R_{U} >}_{SOP} = \frac{1}{2 K} \underset{k}{Σ} (R_{U}^{(k)} + R_{U}^{'' (k)}) - - - (18 e)

Perhaps, when the coupling mechanism ratio can be ignored in the range of wavelengths of regulation with wavelength change,＜R _L _SOPWith＜R _U _SOPCan be replaced by:

{< R >}_{SOP; λ} = \frac{1}{4 K} \underset{k}{Σ} (R_{L}^{(k)} + {R^{''}}_{L}^{(k)} + R_{U}^{(k)} + {R^{''}}_{U}^{(k)}) - - - (18 f)

At this, do not require automatic calibration process, that is, can omit above-mentioned with power P _yMultiply by this process of relative gain that records in advance.

(iii) for the embodiment (single OPTICAL SENSORS) of Figure 1B, available power only has P _x(obtaining) from OPTICAL SENSORS 22A.This standardization power is as obtaining in (19d), but not at first rated output x to the ratio of power y, promptly

T_{L}^{(k)} = u_{o} \frac{{Px}_{L}^{(k)}}{{< P_{L} >}_{SOP}}

{T^{''}}_{L}^{(k)} = u_{o} \frac{{Px}^{''}_{L}^{(k)}}{{< P_{L} >}_{SOP}}

T_{U}^{(k)} = u_{o} \frac{{Px}_{U}^{(k)}}{{< P_{U} >}_{SOP}}

{T^{''}}_{U}^{(k)} = u_{o} \frac{{Px}^{''}_{U}^{(k)}}{{< P_{U} >}_{SOP}} - - - (18 h)

Wherein average power is defined as,

{< P_{L} >}_{SOP} = \frac{1}{2 K} \underset{k}{Σ} ({Px}_{L}^{(k)} + {Px}^{''}_{L}^{(k)})

{< P_{U} >}_{SOP} = \frac{1}{2 K} \underset{k}{Σ} ({Px}_{U}^{(k)} + {Px}^{''}_{U}^{(k)}) - - - (18 i)

At this, suppose that emissive power is stable in the power measurement process.

(iv) for Fig. 1 H embodiment---it has two OPTICAL SENSORS, and combine a coupling mechanism that is positioned at after the analyzer, obtain two power: P from

OPTICAL SENSORS

22B and 22C respectively _xPower and P _x" power.Standardization power (transmitance) is now obtained to be,

T_{L}^{(k)} = u_{o} \frac{{Px}_{L}^{(k)}}{{< {Px}_{L} >}_{SOP}}

{T^{''}}_{L}^{(k)} = u_{o} \frac{{Px}^{''}_{L}^{(k)}}{{< {Px}^{''}_{L} >}_{SOP}}

T_{U}^{(k)} = u_{o} \frac{{Px}_{U}^{(k)}}{{< {Px}_{U} >}_{SOP}}

{T^{''}}_{U}^{(k)} = u_{o} \frac{{Px}^{''}_{U}^{(k)}}{{< {Px}^{''}_{U} >}_{SOP}} - - - (18 j)

Wherein average power is defined as,

{< {Px}_{L} >}_{SOP} = \frac{1}{K} \underset{k}{Σ} {Px}_{L}^{(k)}

{< {Px}^{''}_{L} >}_{SOP} = \frac{1}{K} \underset{k}{Σ} {Px}^{''}_{L}^{(k)}

{< {Px}_{U} >}_{SOP} = \frac{1}{K} \underset{k}{Σ} {Px}_{U}^{(k)}

{< {Px}^{''}_{U} >}_{SOP} = \frac{1}{K} \underset{k}{Σ} {Px}^{''}_{U}^{(k)} - - - (18 k)

At this, do not want automatic calibration process yet.Notice that this embodiment has such advantage: compared with other embodiments, only require obtaining (both data samplings) time of half.

Notice that for (iii) above-mentioned and (iv) standardization, the power during the measurement must be stable.And if for all wavelengths in the range of wavelengths of regulation, power is constant, so can on the SOP or on the wavelength or SOP and wavelength right on the two＜ _SOPAverage.

Basically, if all these relations are all effective for all scenario---applied enough SOP of input and output at random and disturbed partially, provided correct DGD value, might obtain the relation of DGD and mid point wavelength then at a specific mid point wavelength place.Therefore, also can calculate average DGD or rms DGD value at given range of wavelengths.

Under other situations, scanning mid point wavelength is---according to the definition of PMD---to consider on wavelength DGD being averaged, obtaining rms DGD value (not being average DGD).On the contrary, as mentioned above, only on wavelength, carry out mean deviation and keep input and output SOP constant simultaneously, require to satisfy described supposition, and require big PMD Δ v product value about this FUT.Identical note is applicable to the equation that after this presents.

3.2 noise variance

Second motivation that sampling repeats track data is, from the variation of light polarization and/or laser optics frequency and/or power (intensity), obtain the ability of the accurate estimation of variance noise, described sampling repeat track under muting situation for SOP and mid point wavelength X _MidEvery kind of setting be identical basically.If this noise variance is known, then it can be deducted.Because these repeat tracks, just can be estimated as follows independently from the variance of polarization noise and/or laser optics frequency and/or power noise and/or any other noise or the like:

σ_{noise}^{2} = {(\frac{1}{σ_{20}})}^{2} {< (T_{L} - {T^{''}}_{L}) (T_{U} - {T^{''}}_{U}) >}_{SOP; λ} - - - (19)

Wherein

σ_{20}^{2} = 1 / 12

It should be noted that this " noise " variance can be from the input and output SOP of random variation, and/or the instability of laser optics frequency and intensity, or any other noise source.

In order to obtain reliable measurement result, this variance noise---for example from polarization variations and similar other effects, such as the instability of laser optics frequency and intensity---should be than mean square deviation (the sub-chapters and sections 3.4 of face as follows) little several percentage points (for example less than 2%).

It should be noted that and above-mentionedly on average can average on the two on the SOP or at SOP and wavelength.

3.3 relative variance

Relative variance, the causing of the unpolarized ASE light of the optical amplifier in for example main origin self-test link (or any other depolarization effect), as be used in equation (10) and (11), this be calculated as four available estimations on average, that is,

{σ^{'}}_{r}^{2} = {(\frac{1}{σ_{20}})}^{2} [\frac{δ (T_{L}) + δ ({T^{''}}_{L}) + δ (T_{U}) + δ ({T^{''}}_{U})}{4}] - - - (20 a)

Wherein

σ_{20}^{2} = 1 / 12,

And function " δ " is defined as,

δ (T_{L}) = [{< {T_{L}}^{2} >}_{SOP; λ} - {< T_{L} >}_{SOP; λ}^{2}]

δ (T_{L}^{''}) = [{< {T_{L}}^{'' 2} >}_{SOP; λ} - {< T_{L}^{''} >}_{SOP; λ}^{2}]

δ (T_{U}) = [{< {T_{U}}^{2} >}_{SOP; λ} - {< T_{U} >}_{SOP; λ}^{2}]

δ (T_{U}^{''}) = [{< {T_{U}}^{'' 2} >}_{SOP; λ} - {< T_{U}^{''} >}_{SOP; λ}^{2}]

Alternatively, also can be via polarized component s _pCalculate relative variance, for example

{σ^{'}}_{r}^{2} = {(\frac{1}{σ_{s 0}})}^{2} [\frac{{< s_{pL}^{2} >}_{SOP; λ} + {< s_{pL}^{'' 2} >}_{SOP; λ} + {< s_{pU}^{2} >}_{SOP; λ} + {< s_{pU}^{'' 2} >}_{SOP; λ}}{4}] - - - (20 b)

Wherein

σ_{s 0}^{2} = 1 / 3,

And s _pFor

s _pL＝2T _L-1 s″ _pL＝2T _L″-1

s _pU＝2T _U-1 s″ _pU＝2T _U″-1

But notice that the relative variance of calculating from equation (20b) can not be applied to any above-mentioned or following and " relative power " the relevant calculating that is used to extract DGD or PMD, that is, the power that records must suitably standardization of quilt.

It should be noted that above-mentioned equation is effective under the situation that I-SOP and A-SOP from input and output Polarization Controller one or both of are evenly distributed on Poincare sphere.Can only on SOP, average, or average on the two at SOP and wavelength.

Then, noise variance (equation 19) is deducted from first of relative variance (equation 20a) is estimated in calculating, thereby final relative variance is as follows,

σ_{r}^{2} = {σ^{'}}_{r}^{2} - σ_{noise}^{2} - - - (21)

3.4 mean square deviation

The calculating here is different from the simply all sides in equation (10) and (11), and in equation (10) and (11) all just is for for the purpose of more clear and do not consider noise.On the contrary, at λ _LAnd λ _UThe product quilt that repeats difference between the standardization power at place is on average as follows,

{< Δ T^{2} >}_{SOP; λ} = {< (T_{U} - T_{L}) \cdot ({T^{''}}_{U} - {T^{''}}_{L}) >}_{SOP; λ} = \frac{1}{K} \underset{k}{Σ} (T_{U}^{(k)} - T_{L}^{(k)}) \cdot ({T^{''}}_{U}^{(k)} - {T^{''}}_{L}^{(k)}) - - - (22)

In the mathematical term of routine, equation (22) can be called as these second order node squares that repeats difference.Like this, noise on average arrives zero, and does not need by " correction ", because the noise that is superimposed upon on the given trace is also uncorrelated with the noise on being superimposed upon corresponding repetition power.Here it is is used to obtain first motivation of repeating data.

Note, in above-mentioned equation,＜ _{SOP; λ}Can refer on the SOP or on the mid point wavelength or on the two, asking average, that is, organize power to another, change SOP or wavelength or both from one group of power.

3.5DGD or the calculating of PMD value

Then, calculate DGD or rms DGD (being PMD) is as follows according to the inverse sine formula,

DGD (v) = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{\frac{{< ΔT {(v)}^{2} >}_{SOP}}{σ_{r}^{2} (v)}}) - - - (23)

Wherein＜ _SOPOnly refer to average on SOP.

PMD = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{\frac{{< Δ T^{2} >}_{SOP; λ}}{σ_{r}^{2}}}) - - - (24)

Wherein＜ _{SOP; λ}Refer at average on the two of SOP and wavelength, theoretical constant

α_{ds} = \sqrt{\frac{9}{2}} .

It will be appreciated that the inverse sine formula in equation (23) and (24) is not only feasible formula.The purpose of using this formula is in order to obtain not offset result, even using big relatively stride, such as making PMD δ v～0.2, situation under, also can obtain not offset result, and not introduce significant error; This is in order to make signal to noise ratio (S/N ratio) maximization, maximizes thereby make between the dynamic area of this instrument.Though can be applicable to the stride of any size, if do not consider to make between dynamic area to maximize, then can select little stride, following in this case simpler differential formulas is effective:

DGD (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< ΔT {(v)}^{2} >}_{SOP}}{σ_{r}^{2} (v)}} - - - (23 a)

PMD = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< Δ T^{2} >}_{SOP; λ}}{σ_{r}^{2}}} - - - (24 a)

Here be not the better or advantageous particularlies of these formula of hint, and only because, if stride is much smaller, PMD δ v＜0.01 that promptly satisfies condition, they also can be convenient to use.

It should be noted that under ideal conditions---wherein not from the ASE of optical amplifier and " depolarizing " effect and other " noise " of polarization, frequency and intensity or the like,

σ_{r}^{2} = 1,

Above-mentioned equation (23) and (24) by abbreviation are,

DGD (v) = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{{< ΔT (v^{2}) >}_{SOP}}) - - - (25)

PMD = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{{< Δ T^{2} >}_{SOP; λ}}) - - - (26)

Their simpler differential formulas is,

DGD (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{{< Δ T {(v)}^{2} >}_{SOP}} - - - (25 a)

PMD = \frac{α_{ds}}{πδv} \cdot \sqrt{{< Δ T^{2} >}_{SOP; λ}} - - - (26 a)

It should be noted that in above-mentioned equation,＜ _{SOP; λ}Can refer to averaging on the two on the wavelength or at SOP and wavelength, that is, organize power to another, change SOP and wavelength from one group of power.

Note, can by to from the DGD of a lot of different mid point wavelength (v), on the range of wavelengths of regulation, ask average, calculate average DGD or rms DGD, such as

RMS DGD = \sqrt{{< {DGD}^{2} >}_{λ}} - - - (27)

mean?DGD＝<DGD> _λ (28)

Shown in equation (23) and (24), if DGD (v) calculates with PMD and relates to the relative variance σ that uses standardization power (T) _r ²(v), this standardization power can be normalized between 0 and 1 so.In other words, above-mentioned some step that is used to obtain the standardisation process of standardization power can be skipped.

For example, for the embodiment (two OPTICAL SENSORS connect same coupling mechanism) of Fig. 1 C, relative power (P _R) can be simply from track P _xTo track P _yRatio obtain, for,

P_{RL}^{(k)} = \frac{{Px}_{L}^{(k)}}{{Py}_{L}^{(k)}}

{P^{''}}_{RL}^{(k)} = \frac{{Px}^{''}_{L}^{(k)}}{{Py}^{''}_{L}^{(k)}}

P_{RU}^{(k)} = \frac{{Px}_{U}^{(k)}}{{Py}_{U}^{(k)}}

{P^{''}}_{RU}^{(k)} = \frac{{Px}^{''}_{U}^{(k)}}{{Py}^{''}_{U}^{(k)}} - - - (29)

For the embodiment (two OPTICAL SENSORS connect same coupling mechanism) among the embodiment among Fig. 1 D (two OPTICAL SENSORS connect same PBS) and Fig. 1 C, anyly on SOP and/or wavelength, ask average process, for obtaining relative power (P with reference to constant with in order to obtain standardization power _R) process, can be omitted (skipping).Then, can use following inverse sine formula to calculate DGD and PMD, for,

DGD (v) = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{\frac{{< Δ P_{R} {(v)}^{2} >}_{SOP}}{σ_{R}^{2} (v)}}) - - - (30)

Wherein＜ _SOPOnly refer to average on SOP.

PMD = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{\frac{{< Δ P_{R}^{} >}_{SOP; λ}}{σ_{R}^{2}}}) - - - (31)

Wherein＜ _{SOP; λ}Refer to average on the two of SOP and wavelength.At this, equal side＜Δ P _R ² _{SOP: λ}Can be obtained as follows,

{< Δ P_{R}^{} >}_{SOP; λ} = {< (P_{RU} - P_{RL}) \cdot ({P^{''}}_{RU} - {P^{''}}_{RL}) >}_{SOP; λ} = \frac{1}{K} \underset{k}{Σ} (P_{RU}^{(k)} - P_{RL}^{(k)}) \cdot ({P^{''}}_{RU}^{(k)} - {P^{''}}_{RL}^{(k)}) - - - (32 a)

Relative variance σ _R ²Be calculated as the average of four available estimations at this, that is,

σ_{R}^{2} = {(\frac{1}{σ_{20}})}^{2} [\frac{δ (P_{RL}) + δ ({P^{''}}_{RL}) + δ (P_{RU}) + δ ({P^{''}}_{RU})}{4}] - - - (32 b)

Wherein

σ_{20}^{2} = 1 / 12,

And function " δ " is defined as,

δ (P_{RL}) = [{< {P_{RL}}^{2} >}_{SOP; λ} - {< P_{RL} >}_{SOP; λ}^{2}]

δ (P_{RL}^{''}) = [{< {P_{RL}}^{'' 2} >}_{SOP; λ} - {< P_{RL}^{''} >}_{SOP; λ}^{2}]

δ (P_{RU}) = [{< {P_{RU}}^{2} >}_{SOP; λ} - {< P_{RU} >}_{SOP; λ}^{2}]

δ (P_{RU}^{''}) = [{< {P_{RU}}^{'' 2} >}_{SOP; λ} - {< P_{RU}^{''} >}_{SOP; λ}^{2}]

Note,＜ _{SOP; λ}Can refer to average on the SOP or average on the wavelength or at it average on the two, that is, organize power to another from one group of power, change SOP and wavelength.

If select little stride, then the inverse sine formula in equation (30) and (31) can be written as simpler differential formulas, for,

DGD (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< Δ P_{R} {(v)}^{2} >}_{SOP}}{σ_{R}^{2} (v)}} - - - (30 a)

PMD = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< Δ P_{R}^{} >}_{SOP; λ}}{σ_{R}^{2}}} - - - (31 a)

For tunable optical source have big relatively live width and the test be the situation of high PMD optical fiber, in formula, can further use live width " modifying factor ", with DGD or the PMD value that draws FUT more accurately.

It should be noted that above-mentioned calculate both-end the PMD DGD forward or the PMD that measure, be DGD or the PMD value of FUT.

Data processing and calculating at single-ended overall PMD measurement

1. data structure

Coming each back reflective luminous power of the local reflection (such as Fresnel reflection) of comfortable FUT far-end---it is described wavelength of method of operating and the (I-SOP that measures at single-ended overall PMD, A-SOP) right given a setting down obtains---constituted a primitive, that is, data are made of a performance number.Next data unit is one four power packages (i.e. four data unit), embodiment for Fig. 2 C and Fig. 2 G, be four back reflective power of two covers, wherein two back reflective power obtain from

OPTICAL SENSORS

22B and 22C simultaneously, all all I/O-SOP controller 14 be provided with given (I-SOP is A-SOP) to obtaining down.These two covers, four power packages that form the k group preferably obtain (time flows to right from a left side) with following sequence:

Wherein mark x and y and refer to the measurement power that (or in different slightly time) simultaneously obtains from

OPTICAL SENSORS

22B and 22C respectively, λ _U ^(k)-λ _L ^(k)Equal stride δ λ, the mid point wavelength is defined as

λ_{k} = (λ_{U}^{(k)} + λ_{L}^{(k)}) / 2,

The indication of two left-falling stroke symbol repeats power.

Finally, obtaining after all measure power, they will be stored in its conceptual data in the data file and be drawn into matrix in the following equation (31), will be with reference to this matrix in chapters and sections subsequently.This matrix comprises K group, and each group is formed (when using two OPTICAL SENSORS, two covers, four luminous power groups being arranged) by four back reflective luminous powers:

2. the automatic calibration of relative gain

For the preferred embodiment of the use polarization beam apparatus (PBS) of Fig. 2, before proceeding any further calculating, be necessary to carry out the calibration process of the relative gain of two detecting device 22B and 22C.Its calibration process will be described below.For other embodiments, then do not carry out this process.

This calibrating principle is based on the following fact: the SOP that is used to generate enough big quantity when I/O-SOP scrambler 14 is when evenly covering Poincare sphere basically, the ratio that from the average power of the back reflective light of the far-end (or other positions) of FUT 18 will be 2: 1 comes out from two ports of PBS, the port that this higher power is connected corresponding to detecting device 22B, the port that this lower power is connected corresponding to detecting device 22C.Therefore, for viewed detecting device power, any viewed and departing from of this 2: 1 ratios can be detected and include in following consideration.

Data obtain finish after, K four back reflective luminous powers that obtain from two OPTICAL SENSORS are stored, and, come J=4K altogether the power (data) of self-detector 22B that is, what come self-detector 22C also is the individual track of J=4K, describes as matrix (31).J power (j=0,1... (J-1)) from 22B and 22C is the following Px that is claimed respectively _jAnd Py _jIf the overall loss in two arms of PBS equates, and the gain of two OPTICAL SENSORS and related electron device also equates, so power P _yAnd P _xRatio after asking on average on all J time will be

\frac{< Px >}{< Py >} &equiv; \frac{\underset{j}{Σ} {Px}_{j}}{\underset{j}{Σ} {Py}_{j}} = 2

In practice, from the average of the power that records and ratio that obtain may be not equal to 2 because the loss in two arms of PBS may be different, and " effectively " of two OPTICAL SENSORS to gain may be different; Described " effectively " gain comprises the overall gain of photodiode response rate and following electron device---amplifier and sample circuit---.(note, needn't determine these individual gains respectively.) therefore, before proceeding all the other calculating, from all J power of OPTICAL SENSORS 22C acquisition, i.e. all Py _j, it is as follows all to make multiplication:

Py _j≡g _RoundTrip·Py _j

Wherein

g_{RoundTrip} = \frac{1}{2} \frac{< Px >}{< Py >} = \frac{\underset{j}{Σ} {Px}_{j}}{\underset{j}{Σ} {Py}_{j}}

In practice, for the centre wavelength of those relatively closely alternate (for example less than 20nm), the relative wavelength dependency of member, detecting device or the like may be insignificant, and this calibration process is measured sequence to each single-ended PMD only needs execution once both can.Otherwise, may carry out such calibration in each central wavelength, prolonged the overall measurement time of this measurement sequence thus.

The result of this calibration is, promptly at all P _yPower (data) all multiply by after the relative gain that records as described above, and data processor 34 can normalized back reflective luminous power.More accurately, under the situation of the embodiment of the use PBS of Fig. 2, standardization power is obtained as follows: with sample and the average signal P that comes self-detector 22B _x, or come the signal P of self-detector 22C _y, or (and being preferably) multiply by poor (Px-Py)/2 behind the weight coefficient or (Py-Px)/2 (will describe in detail at next chapters and sections), or any poor (1+w) that multiply by behind the weight coefficient ^-1(Px-wPy) (wherein w is a weighting factor), divided by come self-

detector

It should be noted that other calibration stepss also are feasible.For example, a kind of potential substituting collimation technique is: use fiber coupler (beam splitter) or inner reflector as internal reference, sending to two different detecting devices from the luminous power of the predetermined amount (number percent) of the OTDR light of being launched.

3. calculate

Power is processed, with the PMD value that obtains as will describe now.It should be noted that hereinafter all this symbol points to the matrix " data " in the equation (33), mark x and y refer to respectively from the back reflective luminous power of

OPTICAL SENSORS

22B and 22C acquisition.

3.1 standardization power

Standardization power (being transmissivity) after this is noted as T, and it calculates according to the difference of each embodiment and difference.

(i) for the embodiment (have two OPTICAL SENSORS and connect same PBS) of Fig. 2, the standardisation process of the embodiment of describing in the calculating of standardization power and the previous related Sections, be used for Fig. 1 D of measuring at both-end PMD (having two OPTICAL SENSORS to connect same PBS) is identical.But note, before being used in this standardisation process, different P _yPower must be multiply by the relative gain g that records from single-ended measurement in advance _RoundTrip, as indicated in the description of automatic calibration process.

(ii) for the embodiment (having two OPTICAL SENSORS to connect same coupling mechanism) of Fig. 2 D, the calculating of standardization power also with previous related Sections in the standardisation process of the embodiment of measuring at both-end PMD (having two OPTICAL SENSORS to connect same coupling mechanism) that describe, that be used for Fig. 1 C identical.But note, in this standardisation process, different reference average u has been used in single-ended measurement _o=2/3.

At this, do not require automatic calibration process, promptly can omit above-mentioned with power P _yMultiply by this process of relative gain that records in advance.

(iii) for the embodiment (single OPTICAL SENSORS) of Fig. 2 C, equally, the standardisation process of the embodiment of describing in the calculating of standardization power and the previous related Sections, be used for Figure 1B of measuring at both-end PMD (two OPTICAL SENSORS connect same coupling mechanism) is identical, and also must use with reference to average u single-ended measurement in this standardisation process _o=2/3.

At this, suppose the λ that is transmitted among the FUT _U ^(k)And λ _L ^(k)The power at place is approximately uniform.

It should be noted that in above-mentioned equation,＜ _{SOP; λ}Can refer on the I-SOP or on the A-SOP or on the mid point wavelength, average on whole threes ideally, promptly organize power to another from one group of power, correspondingly change (I-SOP, A-SOP) to the mid point wavelength.Even all these relations are all effective basically for all scenario---only apply polarization and disturb partially, provided the correct DGD value at a specific mid point wavelength place.Like this, scanning mid point wavelength is---according to the definition of static PMD value---to consider to obtain on wavelength DGD being averaged.On the contrary, as mentioned above, only on wavelength, carry out mean deviation and keep simultaneously that (I-SOP A-SOP) to constant, require to satisfy described supposition about this FUT, and requires big product value PMD Δ v.Identical note is applicable to the equation that after this presents.

3.2 mean square deviation

The calculating here is different from the simply all sides in equation (1), (2), (12) and (13), and in equation (1), (2), (12) and (13) simply all just is for for the purpose of more clear and do not consider noise.On the contrary, at λ _LAnd λ _UThe product quilt that repeats difference between the standardization track at place is on average as follows,

{< Δ T^{2} >}_{SOP; λ} = {< (T_{U} - T_{L}) \cdot ({T^{''}}_{U} - {T^{''}}_{L}) >}_{SOP; λ} = \frac{1}{K} \underset{k}{Σ} (T_{U}^{(k)} - T_{L}^{(k)}) \cdot ({T^{''}}_{U}^{(k)} - {T^{''}}_{L}^{(k)}) - - - (2 2^{,})

Notice that equation (22 ') is identical with equation (22).In the mathematical term of routine, equation (22 ') can be called as the second order node square that repeats difference.Like this, noise can be averaged in to zero, rather than by " correction ", because the noise that is superimposed upon on the given trace is uncorrelated with the noise on being superimposed upon corresponding repetition track.This is first motivation of obtaining the repetition track.

3.3PMD the calculating of value

Then, according to following inverse sine formula, PMD directly is calculated as,

PMD = α_{rt} \frac{1}{πδv} \arcsin (α_{ds} \sqrt{{< ΔT {(v)}^{2} >}_{SOP; λ}}) - - - (34)

Wherein, come and go the factor

α_{rt} = \sqrt{\frac{3}{8}} .

Theoretical constant

α_{ds} = \sqrt{\frac{15}{4}}

Be used to control the situation of the SOP of input and output light for wherein identical polarization state controller (scrambler), as the situation of Fig. 2,2C-G, effectively.

It will be appreciated that the inverse sine formula in the equation (34) is not only feasible formula.The purpose of using this formula is to obtain not offset result, even using big relatively frequency step, under the situation such as PMD δ v～0.15, also can obtain not offset result, and not introduce significant error; This is in order to make signal to noise ratio (S/N ratio) maximization, maximizes thereby make between the dynamic area of this instrument.If do not consider to make between dynamic area to maximize, or keep the overall measurement time reasonable, then can select much smaller frequency step, and the simpler differential formulas below using:

PMD = α_{rt} \cdot \frac{α_{ds}}{πδv} \cdot \sqrt{{< ΔT {(v)}^{2} >}_{SOP; λ}} - - - (34 a)

This is not the better or advantageous particularlies of these formula of hint, and only because, if frequency step is much smaller, PMD δ v＜0.01 that promptly satisfies condition, they also can be convenient to use.

It should be noted that from equation (34) and (34a) calculate PMD forward be PMD or the rmsDGD of FUT.

Also it should be noted that for a given range of wavelengths, can pass through the DGD under the much different wavelength _RoundTrip(v) carry out root mean square or on average try to achieve coming and going rms DGD or average DGD, and at the DGD at each setted wavelength place _RoundTrip(v) can calculate from the inverse sine formula, for,

{DGD}_{RoundTrip} (v) = \frac{1}{πδv} \arcsin (α_{ds} \sqrt{{< ΔT {(v)}^{2} >}_{SOP}}) . - - - (35)

Or use simpler differential formulas as follows,

{DGD}_{RoundTrip} (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{{< ΔT {(v)}^{2} >}_{SOP}} . - - - (35 a)

Wherein standardization power (T) obtains at each setted wavelength.

Also can pass through simply to rms DGD _RoundTripMultiply by the round factor respectively

Obtain rms DGD and average DGD (forward) with 2/ π, wherein, the DGD that can record from many different mid point wavelength _RoundTrip(v) obtain rms DGD _RoundTripOr equal DGD _RoundTrip---by to from equation (35) or DGD (35a) _RoundTrip(v) on the range of wavelengths of a regulation, carry out root mean square or average, for example

rms {DGD}_{RoundTrip} = \sqrt{{< {DGD}_{RoundTrip}^{2} >}_{λ}}

With average DGD _RoundTrip=＜DGD _RoundTrip _λ

Also it should be noted that and above-mentionedly be used to use standardization power (the normally standardization power between 0 to 1) DGD that draws and the calculation equation of PMD can be replaced by additive method.For example, only calculate relative power from the power that records, " normalization factor " that be used in equation (34) and (35) then can also be taken advantage of on mean square deviation because of the same factor when mean square deviation is removed by cancellation, thereby obtains correct DGD or PMD value.

It should be noted that the above-mentioned equation that is used to calculate DGD or PMD has theoretical constant

α_{ds} = \sqrt{\frac{15}{4}} .

This theoretical constant value is effective for wherein having the situation---such as situation of Fig. 2,2C-G---that identical common polarization state controller (scrambler) is used as input and output light SOP control.Yet, when two discrete independent input and output polarization state controllers (scrambler) with just be placed on detecting device before polarizer or PBS when using jointly, for example shown in Fig. 2 G, must use different theoretical constants

α_{ds} = \sqrt{\frac{9}{2}} .

(notice that this theoretical constant theoretical constant described with above-mentioned related Sections, that be used for both-end PMD measurement equation is identical.)

When having big relatively live width and testing the situation of high PMD optical fiber, in equation (8-11), may need to use live width " modifying factor ", thereby obtain this FUT PMD value accurately for tunable light-pulse generator.

Data processing and calculating at single-ended accumulation PMD measurement

1. data structure

Each OTDR track---its be measure at single-ended accumulation PMD the wavelength described in the method for operating of (the accumulation PMD that is also referred to as based on single-ended POTDR measures) and (I-SOP, A-SOP) right one givenly is provided with down acquisition---constituted primitive.A track is made of N performance number, and this N performance number is corresponding to N value z of distance z _n, n=0... (N-1) wherein.

Next bigger data unit is four trajectory set, embodiment for Fig. 3 and Fig. 3 B is two covers, four trajectory set, wherein two tracks obtain (perhaps from

OPTICAL SENSORS

22B and 22C simultaneously, under the situation of an applied optics switch and a detecting device, one after the other obtain two tracks), all all I/O-SOP controller 14 be provided with given (I-SOP is A-SOP) to obtaining down.These two covers, four trajectory set that form k group preferably obtain (time flows to right from a left side) with following sequence, wherein mark x and y and refer to respectively simultaneously the track that obtains from

OPTICAL SENSORS

22B and 22C, λ _U ^(k)-λ _L ^(k)Equal wavelength stride δ λ, the mid point wavelength is defined as

λ_{k} = (λ_{U}^{(k)} + λ_{L}^{(k)}) / 2,

The indication of two left-falling stroke symbol repeats power:

Finally, after data are obtained, be stored in conceptual data in the data file and be drawn into matrix in the following equation (36).Later on will be with reference to this matrix.This matrix comprises K group, and each group has four OTDR tracks (when using two OPTICAL SENSORS, being two covers, four trajectory set), and each track is made of N point, this N N value z that puts corresponding to distance z _n, n=0... (N-1) wherein:

The data structure of equation (36) is similar to equation (33), but the data in the equation (36) are the OTDR tracks that change along with distance z, rather than the power from the reflection of FUT far-end in the equation (33).

2. the automatic calibration of relative gain

For the preferred embodiment of Fig. 3, before proceeding any further calculating, be necessary to carry out the following calibration process of the relative gain of two detecting device 22B and 22C.For other embodiments, then do not carry out this process.

This calibration principle is based on following true the setting: when I/O-SOP scrambler 14 is used to generate when enough the SOP of big quantity also evenly covers Poincare sphere basically, average power at the back reflective light on any fragment of FUT 16 will be exported from two ports of PBS with 2: 1 ratio, the port that this higher power is connected corresponding to detecting device 22B, the port that this lower power is connected corresponding to detecting device 22C.Therefore, for viewed detecting device power, any viewed and departing from of this 2: 1 ratios can be quantized and include in consideration, and be as follows.

Data obtain finish after, K four OTDR trajectory set that obtain from two OPTICAL SENSORS are stored, and, come J=4K altogether the track of self-detector 26A that is, what come self-detector 22B also is J=4K track, describes as matrix (36).Below j the track (j=0,1... (J-1)) of 22C and 22B, be called as Px (z) respectively _jAnd Py (z) _jIf the overall loss in two arms of PBS equates, and the gain of two OPTICAL SENSORS and related electron device also equates, so track P _yAnd P _xRatio after asking on average on all J time and on the N of all z value will be

\frac{< Px >}{< Py >} &equiv; \frac{\underset{j}{Σ} \underset{n}{Σ} Px {(z_{n})}_{j}}{\underset{j}{Σ} \underset{n}{Σ} Py {(z_{n})}_{j}} = 2

In practice, ratio that obtain may be not equal to 2 from the track that records is average, because the loss difference in two arms of PBS, and " effectively " of two OPTICAL SENSORS gain also may be different; Described " effectively " gain comprises the overall gain of photodiode response rate and following electron device---amplifier and sample circuit---.(note, needn't determine these individual gains respectively.) therefore, before proceeding all the other calculating, from all J track of OPTICAL SENSORS 22C acquisition, i.e. all Py (z) _j, it is as follows all to make multiplication:

Py(z) _j≡g _RoundTripC·Py(z _n) _j

Wherein

g_{RoundTripC} = \frac{1}{2} \frac{< Px >}{< Py >} = \frac{\underset{j}{Σ} \underset{n}{Σ} Px {(z_{n})}_{j}}{\underset{j}{Σ} \underset{n}{Σ} Py {(z_{n})}_{j}}

In practice, for the mid point wavelength of those relatively closely alternate (for example less than 20nm), the relative wavelength dependency of member, detecting device or the like may be negligible, and this calibration process is measured sequence to each POTDR only to be needed to carry out once.Otherwise, may carry out this calibration in each central wavelength, prolonged the overall measurement time of this measurement sequence thus.

The result of this calibration is, promptly at all P _yTrack all multiply by after the relative gain that records as described above, and data processor 34 can normalized OTDR track.Be that under the situation of the embodiment of Fig. 1, the standardization track is obtained as follows more accurately: with sample and the average signal P that comes self-detector 22B _x, or come the signal P of self-detector 22C _y, preferably coming poor between the sampled signal of self-

detector

22B and 22C, (Px-Py)/2 or (Py-Px)/2 (will describe in detail) at next chapters and sections, or any poor (1+w) that multiply by behind the weight coefficient ^-1(Px-wPy), divided by come self-

detector

22B and 22C the two sample and average signal sum (Px+Py), should and expression incide total back reflective power on this PBS, promptly do not select the general power of specific polarization component.

To describe computing method below in detail to the standardization OTDR track of all preferred embodiments.

3. pointwise is calculated

The processing that will describe the OTDR track now is to obtain the PMD value of accumulation.It should be noted that at each some z along FUT 18 _nThe PMD at place _nCalculating be independent of any other n.Each all derive from those (I-SOP, A-SOP) to last and/or to asking average on the wavelength average that come.Like this, in following calculating, it is inappropriate using subscript n; Must understand simply, this calculating is repeated in an identical manner for each some n, or in other words, in each distance z _nThe place all is effective.At all hereinafter, these symbols point to corresponding with matrix " data " symbol in the equation (36).Should emphasize that also mark x and y refer to respectively from the track of

OPTICAL SENSORS

22B and 22C acquisition.

3.1 standardization track

After this standardization track is noted as T (z), and it calculates according to the difference of each embodiment and different.

(i) for the embodiment (have two OPTICAL SENSORS and connect same PBS) of Fig. 3, standardization OTDR track is calculated as follows:

T_{L}^{(k)} = \frac{{Px}_{L}^{(k)}}{{Px}_{L}^{(k)} + {Py}_{L}^{(k)}}

{T^{''}}_{L}^{(k)} = \frac{{Px}^{''}_{L}^{(k)}}{{Px}^{''}_{L}^{(k)} + {Py}^{''}_{L}^{(k)}}

T_{U}^{(k)} = \frac{{Px}_{U}^{(k)}}{{Px}_{U}^{(k)} + {Py}_{U}^{(k)}}

{T^{''}}_{U}^{(k)} = \frac{{Px}^{''}_{U}^{(k)}}{{Px}^{''}_{U}^{(k)} + {Py}^{''}_{U}^{(k)}}

Or

T_{L}^{(k)} = \frac{1}{2} \cdot \frac{{Px}_{L}^{(k)} - {Py}_{L}^{(k)}}{{Px}_{L}^{(k)} + {Py}_{L}^{(k)}}

{T^{''}}_{L}^{(k)} = \frac{1}{2} \cdot \frac{{Px}^{''}_{L}^{(k)} - {Py}^{''}_{L}^{(k)}}{{Px}^{''}_{L}^{(k)} + {Py}^{''}_{L}^{(k)}}

T_{U}^{(k)} = \frac{1}{2} \cdot \frac{{Px}_{U}^{(k)} - {Py}_{U}^{(k)}}{{Px}_{U}^{(k)} + {Py}_{U}^{(k)}}

{T^{''}}_{U}^{(k)} = \frac{1}{2} \cdot \frac{{Px}^{''}_{U}^{(k)} - {Py}^{''}_{U}^{(k)}}{{Px}^{''}_{U}^{(k)} + {Py}^{''}_{U}^{(k)}} - - - (37 a)

Wherein it will be appreciated that and be used to equation (37) before, different P _yTrack multiply by the relative gain g that records in advance _RoundTripC, indicated as what in automatic calibration process, describe.

(ii) for the embodiment (two OPTICAL SENSORS connect same coupling mechanism) of Fig. 3 B, track P _xTo P _yRatio at first be calculated as,

R_{L}^{(k)} = \frac{{Px}_{L}^{(k)}}{{Py}_{L}^{(k)}}

{R^{''}}_{L}^{(k)} = \frac{{Px}^{''}_{L}^{(k)}}{{Py}^{''}_{L}^{(k)}}

R_{U}^{(k)} = \frac{{Px}_{U}^{(k)}}{{Py}_{U}^{(k)}}

{R^{''}}_{U}^{(k)} = \frac{{Px}^{''}_{U}^{(k)}}{{Py}^{''}_{U}^{(k)}} - - - (37 b)

Then, above-mentioned ratio is by standardization---and as follows with respect to its on average and in addition standardization on K group,

T_{L}^{(k)} = u_{o} \frac{R_{L}^{(k)}}{{< R >}_{SOP; λ}}

{T^{''}}_{L}^{(k)} = u_{o} \frac{{R^{''}}_{L}^{(k)}}{{< R >}_{SOP; λ}}

T_{U}^{(k)} = u_{o} \frac{R_{U}^{(k)}}{{< R >}_{SOP; λ}}

{T^{''}}_{U}^{(k)} = u_{o} \frac{{R^{''}}_{U}^{(k)}}{{< R >}_{SOP; λ}} - - - (37 c)

Wherein with reference to average u _o=2/3---the power measurement of supposition by being parallel to the light polarization state of analyzer axle to one, R is defined as from its average ratio,

{< R >}_{SOP; λ} = \frac{1}{4 K} \underset{k}{Σ} (R_{L}^{(k)} + {R^{''}}_{L}^{(k)} + R_{U}^{(k)} + {R^{''}}_{U}^{(k)}), - - - (37 d)

At this, do not require automatic calibration process, promptly can omit above-mentioned with track P _yMultiply by this process of relative gain that records in advance.

(iii) for the embodiment (single light sensation detector) of Fig. 3 A, only available track is P _xTrack (it obtains from light sensation detector 22).Obtaining of this standardization track as (5c), but need not at first calculate the ratio of track x to track y, that is,

T_{L}^{(k)} = u_{o} \frac{{Px}_{L}^{(k)}}{{< P >}_{SOP; λ}}

{T^{''}}_{L}^{(k)} = u_{o} \frac{{Px}^{''}_{L}^{(k)}}{{< P >}_{SOP; λ}}

T_{U}^{(k)} = u_{o} \frac{{Px}_{U}^{(k)}}{{< P >}_{SOP; λ}}

{T^{''}}_{U}^{(k)} = u_{o} \frac{{Px}^{''}_{U}^{(k)}}{{< P >}_{SOP; λ}} - - - (37 e)

Wherein mean trajectory is defined as,

{< P >}_{SOP; λ} = \frac{1}{4 K} \underset{k}{Σ} ({Px}_{L}^{(k)} + {Px}^{''}_{L}^{(k)} + {Px}_{U}^{(k)} + {Px}^{''}_{U}^{(k)}) - - - (37 f)

It should be noted that in above-mentioned equation,＜ _{SOP; λ}Can refer at I-SOP _kGo up or at A-SOP _kGo up or on the mid point wavelength, average on whole threes ideally, that is, organize track to another from one group of track, correspondingly change I-SOP, A-SOP and wavelength.Even all these relations are substantially all effective for all scenario---only apply I/O-SOP and disturb partially, provided the correct DGD value at a specific mid point wavelength place.Like this, scanning mid point wavelength only is on wavelength DGD being averaged---according to the definition of statistics PMD---considers.On the contrary, as mentioned above, only on wavelength, carry out mean deviation and keep I/O-SOP constant simultaneously, require to satisfy, but also require big product value PMD Δ v about supposition to this FUT.Identical note is applicable to the equation that after this presents.

3.2 relative variance

Relative variance as equation (37b), is calculated as transmission change average of four available estimations at this, that is,

{σ^{'}}_{r}^{2} = {(\frac{1}{σ_{10}})}^{2} [\frac{var (T_{L}) + var ({T^{''}}_{L}) + var (T_{U}) + var ({T^{''}}_{U})}{4}] - - - (38)

Wherein with reference to variance

σ_{10}^{2} = 4 / 45,

And the function of transmission change " var " is defined as,

var (T_{L}) = [{< {T_{L}}^{2} >}_{SOP; λ} - {< T_{L} >}_{SOP; λ}^{2}]

var ({T^{''}}_{L}) = [{< {T^{''}}_{L}^{2} >}_{SOP; λ} - {< {T^{''}}_{L} >}_{SOP; λ}^{2}]

var (T_{U}) = [{< {T_{U}}^{2} >}_{SOP; λ} - {< T_{U} >}_{SOP; λ}^{2}]

var ({T^{''}}_{U}) = [{< {T^{''}}_{U}^{2} >}_{SOP; λ} - {< {T^{''}}_{U} >}_{SOP; λ}^{2}]

3.3 mean square deviation

The calculating here is different from the simply all sides in the equation (3a), and the simply all sides in the equation (3a) are for for the purpose of more clear and do not consider noise.On the contrary, at λ _UAnd λ _LThe product quilt that repeats difference between the standardization track at place is on average as follows,

{< Δ T^{2} >}_{SOP; λ} = {< (T_{U} - T_{L}) \cdot ({T^{''}}_{U} - {T^{''}}_{L}) >}_{SOP; λ} = \frac{1}{K} \underset{k}{Σ} (T_{U}^{(k)} - T_{L}^{(k)}) \cdot ({T^{''}}_{U}^{(k)} - {T^{''}}_{L}^{(k)}) - - - (39)

In the mathematical term of routine, equation (39) can be called as these second order node squares that repeats difference.Like this, noise is averaged in to zero, rather than by " correction ", because the noise that is superimposed upon on the given trace is uncorrelated with the noise on being superimposed upon corresponding repetition track.This is first motivation of obtaining the repetition track.

3.4 noise variance

Second motivation that sampling repeats track is the ability that is to obtain the accurate estimation of variance noise, and it is identical basically that described sampling repeats track every kind of setting for central wavelength lambda and SOP under muting situation.This be because, as calculate in the equation (38) relative variance comprise hypothesis do not have the variance of the track under making an uproar and noise variance the two.Yet if noise variance is known, it can be deducted, because the variance of two independent random variable sums equals these variance sums.But the repetition track has been arranged, and noise just can be estimated independently as follows:

σ_{noise}^{2} = {(\frac{1}{σ_{10}})}^{2} {< (T_{L} - {T^{''}}_{L}) (T_{U} - {T^{''}}_{U}) >}_{SOP; λ} - - - (40)

Then, in the calculating of final relative variance, noise variance (equation 40) can be estimated to deduct (equation 38) from first of relative variance, and is as follows,

σ_{r}^{2} = {σ^{'}}_{r}^{2} - σ_{noise}^{2} - - - (41)

3.5 the calculating of accumulation PMD

Then, according to following inverse sine formula, accumulation PMD is calculated as,

PMD (z) = α_{rt} \frac{1}{πδv} \arcsin (α_{ds} \sqrt{\frac{{< Δ T^{2} (z) >}_{SOP; λ}}{σ_{r}^{2} (z)}}) - - - (42)

Wherein, the round factor is

α_{rt} = \sqrt{\frac{3}{8}} .

Theoretical constant

α_{ds} = \sqrt{\frac{15}{4}}

To wherein using common (identical) common polarization state controller (scrambler) to be used as the situation that input and output light SOP controls, such as the situation of Fig. 3,3A and 3B, effectively.

It will be appreciated that the inverse sine formula in the equation (42) is not only feasible formula.The purpose of using this formula is in order to obtain not offset PMD result of calculation, even using big relatively frequency (wavelength) stride, as makes PMD δ v～0.2, situation under, also can obtain not offset result, and not introduce significant error; This is in order to make signal to noise ratio (S/N ratio) maximization, maximizes thereby make between the dynamic area of this instrument.If do not consider to make between dynamic area to maximize, or keep the overall measurement time reasonable, then can select much smaller frequency step, and the simpler differential formulas below using:

PMD (z) = α_{rt} α_{ds} \frac{1}{πδv} \cdot \sqrt{\frac{{< Δ T^{2} (z) >}_{SOP; λ}}{σ_{r}^{2} (z)}} - - - (43)

This is not the better or advantageous particularlies of these formula of hint, and only because, if stride is much smaller, PMD δ v＜0.01 that promptly satisfies condition, they also can be convenient to use.By corresponding to distance z _nEach some n place repeat above-mentioned calculating, the accumulation PMD curve that acquisition changes along with z from equation (37) to equation (42).

It should be noted that the above-mentioned equation that is used to calculate PMD has theoretical constant

α_{ds} = \sqrt{\frac{15}{4}} .

This theoretical constant value is used as the situation of input and output light SOP control corresponding to wherein identical public (common) polarization state controller (scrambler)---such as to Fig. 3, and the situation of 3A and 3B---effectively.Yet, when two discrete independent input and output polarization state controllers (scrambler) with just be placed on detecting device before polarizer or PBS when using jointly, shown in Fig. 3 C, must use different theoretical constants

α_{ds} = \sqrt{\frac{9}{2}} .

The calculation equation (42) that draws accumulation PMD that also it should be noted that above-mentioned use standardization OTDR track can be replaced by use and the proportional relative OTDR track of standardization OTDR track with (43).

It should be noted that calculate from equation (42) and (43) PMD forward be PMD or the rms DGD of FUT.

3.6 the live width modifying factor of optional application

If effective spectral line width of pulsed laser source is bigger, although then may expect to carry out---being optional---data " aftertreatment " step additional, with the accumulation PMD that considers to record dependence to this laser linewidth.Like this, can be above-mentioned at z to N _nThe accumulation PMD value that the place records, i.e. PMD _n, multiply by a suitable live width dependence modifying factor and obtain accurate PMD value.When this laser linewidth form was the approximate Gaussian form, a kind of suitable modifying factor expression was as follows:

α_{{LW}_{n}} = \frac{1}{\sqrt{1 - {(\frac{{PMD}_{n}}{{PMD}_{sat}})}^{2}}} - - - (43)

PMD wherein _SatBe saturated accumulation PMD value,, were it not for and use the live width modifying factor that is that during the accumulation PMD that records along with cumulative actual PMD trend, it is infinite that saturated accumulation PMD value tends to.It is given:

{PMD}_{sat} = \frac{1}{4 π} \cdot \frac{1}{σ_{vL}} - - - (44)

σ wherein _VLIt is the rms line width of laser spectrum.(note:, be in the overall with at half place of maximal value and the relation of rms width for Gaussian line shape

{Δv}_{L} = \sqrt{8 \cdot \ln (2)} σ_{vL} .

)

Last optional step comprises: according to the modifying factor of equation (44), calculate N PMD value, then, obtain revised PMD value: PMD ' _n, promptly obtain to revise the factor by the PMD that surveys of institute before will revising is on duty,

{PMD}^{'}_{n} = α_{{LW}_{n}} \cdot {PMD}_{n} - - - (45)

For example, were it not for the application modifying factor, from equation (44) and (45) indication, corresponding to for example, the cumulative maximum PMD value of-10% biasing is PMD _Max=0.0817 Δ v _L ^-1A data example of this situation is the overall with Δ v at half place of maximal value _L=2GHz has provided PMD _Sat～93.7ps and PMD _Max～40.8ps.If the value that records equals this predetermined maximal value 40.8ps corresponding to 10% biasing just, so actual PMD in fact is 45.4ps, that is, the value that records is subjected to 10% biasing, as declare.Such possible residual bias level may be acceptable in many rig-site utilization.

Yet, under these identical physical conditions, if modifying factor α _LW=1.11 are applied to equation (45), and then obtaining cumulative actual PMD ' is 45.4ps.

In practice, if that modifying factor becomes is very big, that is, and when (promptly uncorrected) accumulation PMD that directly records too near PMD _SatThen the uncertainty of modifying factor self will increase, because, as noticing from equation (44), in the PMD value that directly records or any little error in the laser linewidth (or effectively the linear uncertainty of laser) can cause modifying factor very unreliable.Yet, if but the maximum permissible value of modifying factor exceed with a predetermined value---this determined then can be measured when modifying factor is employed maximum PMD, should uncertainty keep less.PMD like this, not only _MaxSituation when not revising, and the more important thing is, using under the situation about revising, when actual PMD equals PMD _MaxThe time do not have system biasing, but the uncertainty of little additional average out to zero is only arranged.Use previous examples, modifying factor is set to rational maximal value 1.25, promptly still near unit value, the not maximal value of the actual PMD of biasing that then can be measured is PMD _Max～70ps, the 40.8ps that has-10% biasing when not using the live width modifying factor forms contrast.

Note, whenever product PMD Δ v _LDuring much smaller than unit value, it is nonsensical to use such modifying factor in aftertreatment, in any case because this factor very near unit value.To use modifying factor and be in order to improve, under the situation of the true live width of given laser, can not had the maximum PMD value that biasing ground is measured.

It will be appreciated that equation (44) is applied to the situation of approximate Gaussian shape laser spectrum, as being provided by way of example.Linear for any specific laser that departs from Gaussian line shape basically, can analytical ground or data calculate other formula or relation.Gaussian line shape is wherein to obtain the special case that the simple correlation factor in ground is analyzed formula, also is relevant with practice situation.And this simple analysis formula and be not suitable for that laser is linear arbitrarily.

Be used for the tunable laser source that both-end PMD measures

As mentioned before, it is desirable to have a tunable coherent source, it can be transferred on a lot of mid point wavelength, this wavelength combines the SOP of a lot of input and output simultaneously, its objective is in order to measure the DGD of (as having in the standby DWDM channel of frequency gap of about 35GHz or 70GHz) in any DWDM channel in C or the L band, or obtaining rms or equal DGD value (being the PMD value) exactly, should have enough range of wavelengths to can be used for this measurement this moment.Therefore, it is desirable to, tunable coherent source is tunable on big range of wavelengths.Tunable suitable tunable coherent source is known for those skilled in the art on the interval of hundreds of nanometer, therefore is not described in detail at this.

The tunable optical source of Fig. 7 comprises one based on the fiber optics amplifier---such as SOA---optical fiber ring laser design, wherein shared gain media 102 is used for corresponding at least two different wave lengths (1,2, ..., at least two different cavity N) each (1,2, ..., N).Optical switch 106B is used for the light in different time sections opens or closes these at least two different cavity, and wherein these at least two different wave lengths are selected from simultaneous multichannel tunable fiber 104 by these two different TBF at least.In Fig. 7, at least two polarization adjusters (1,2 ..., if N) be used to regulate the chamber light SOP---the chamber is based on SMF optical fiber cavity.Beam splitter 106A is used to N chamber combined, and coupling mechanism 107 provides light output from laser cavity.Control module 30 ' be used to the regulate centre wavelength of tunable fiber 104, control optical switch " are opened " different laser cavitys sending different wavelength, and the ride gain medium, if for example for the SOA supply of current---SOA is used as gain media.

Fig. 7 A schematically shows an example of a preferred embodiment of so tunable modulated light source (being used among the 12A among Fig. 1 (B-H)), it is designed to send three tight alternate wavelength with sequence fast, and wherein optics isolating switch (chopper) 130 is as optical switch.In a preferred embodiment, the function of TBF 104 can realize with the monolithic diffraction grating, wherein the light path of each of these three laser cavitys incides on the described grating with the different slightly angle in the diffraction plane, and these different slightly angles have been selected as the tight alternate wavelength corresponding to " centre wavelength " of the nominal that centers on this laser.This TBF can tuning " centre wavelength " (as preamble definition) in the one or more bands in the band such as S, C and L or O and E, and concrete inserted wavelength region may depends on the selection of SOA 102 ' and the excessive loss and the wavelength-dependent loss of tunable fiber 104.Preferably, SOA 102 ' is " polarization relies on ", that is to say that it has amplified the input light of specific incident linear polarization best, and amplifies corresponding cross polarization amount indistinctively.The example of such SOA is the Model BOA 1004 that Covega Corporation produces.

Like this, the tunable modulated light source 12A of Fig. 7 A comprises: SOA 102 '; Tunable optical bandpass filter (TBF) 104; Beam-splitting

coupler

106A, 106B and 106C; Optics isolating switch 130; And Three-

port circulator

108A and 108B, it connects with three ring cavity topologys by polarization maintaining optical fibre (PMF).Coupling mechanism 106D combines the light output from

coupling mechanism

106B and 106C.

Control module 30 is coupled to SOA 102 ', isolating switch 122 and TBF 104 by line 120,122 and 124 respectively, it thus supply control signal come optionally the light in the different cavity to be transferred to open or close to close (as hereinafter describing in detail) and regulate wavelength by TBF at different time.

Continuously adjustable TBF has the bandpass filter based on grating of 20 to 40pm tapes wide (FWHM) typically, it is used to tuning laser wavelength exactly, and light (photon) is limited in this little TBF bandwidth, to provide the optical maser wavelength that has narrow linewidth accurately.If use the PMF chamber, then without any need for additional member.If but this chamber then still needs one or two Polarization Controller to regulate the interior polarization state (SOP) of this laser cavity based on for example SMF-28 optical fiber.

The spectral line width of the tunable modulated optical coherent source in the above-mentioned various embodiment can from less than 1GHz to about 4GHz.At least approx it is favourable knowing this live width, facilitates the use the live width modifying factor that preamble is described.

The tunable modulated light source that it will be appreciated that other kinds can be used to substitute the tunable modulated light source that preamble is described.For example, can expect, can utilize external phase modulator to generate the optics sideband at the output terminal of outside cavity gas laser (ECL), distributed bragg reflection laser (DBR) or distributed feedback laser (DFB).

Those skilled in the art will recognize that other substitutes of this tunable modulation coherent source.

Be used for the tunable OTDR that single-ended PMD measures

As mentioned before, it is desirable to use many mid point wavelength X _MidAnd many I-SOP and A-SOP.Therefore, expect that tunable OTDR is tunable on big range of wavelengths.Tunable suitable tunable OTDR is known for those skilled in the art on the interval of hundreds of nanometer, does not therefore describe at this.

Fig. 8 A schematically shows an example of such tunable pulsed laser source 12, and it is disclosed in sequence number of submitting on July 18th, 2006 is that the content of this application is included this instructions by reference in 60/831,448 the total U.S. Provisional Patent Application.Tunable OTDR designs based on annular optical fiber laser, and wherein semi-conductor optical amplifier (SOA) is both as (i) gain medium, again as (ii) external modulator---it also amplifies light pulse when " opening ".(this SOA can be amplified to 17 to 20dBm (outputs) from 3 to 6dBm (inputs) with input optical pulse).

Like this, the tunable pulsed laser source 12 of Fig. 8 A comprises SOA 202, tunable optical bandpass filter (TBF) 204, beam-splitting coupler 206 and four port circulator 208, and they connect with ring topology by polarization maintaining optical fibre (PMF).Coupling mechanism 206 has: first port, and it is connected to SOA 202 via TBF 104; Second port, it is connected to circulator 208 via TMF loop 214; And the 3rd port, it is connected to an end of line of time delay 210, and this end is opposite with that end that with reverberator 212 is termination.Like this, this ring comprises: first path, and it is to amplify the path, extends between circulator 208 and the coupling mechanism 206, and contains SOA 202; And second path, it is a feedback path, extends between coupling mechanism 206 and the circulator 208, is provided by PMF 214.

Coupling mechanism 206 extracts the part of the light in the chambeies, and typically 25 to 50%, and it is sent into line of time delay 210.After device 212 reflections that are reflected, this part light turns back to coupling mechanism 206, and reenters this chamber after the time delay Δ t in the belt travel-time that equals line of time delay 210.Therefore, line of time delay 210 comprises an optical fiber pigtail of polarization maintaining optical fibre, and reverberator 212 comprises a mirror with reflectivity of about 95% in the end of this optical fiber pigtail.Certainly, can use the line of time delay and the reverberator of other suitable form known.

Control module 30 is coupled to SOA 202 and TBF 204 by

line

220 and 222 respectively, and supply control signal optionally opens and closes SOA 202 (as describing in more detail after a while) and regulates the wavelength of TBF 204 whereby.

It should be noted that except the design from Fig. 5 (A) to produce short and high-power light pulse, also can open time ratio and generate long pulse from the much longer time of the time delay of line of time delay 210 by the electric current that makes SOA.

Such tunable pulsed laser source 12 ' can provide high-output power with low cost.About the further details of this tunable pulsed laser source 12 and its operation, the reader can be referring to U.S. Provisional Patent Application No.60/831,448.

The tunable light-pulse generator that it will be appreciated that other kinds also can be used to substitute previously described tunable light-pulse generator.For example, Fig. 8 B is the alternative designs of Fig. 8 A, does not wherein use line of time delay.Design among Fig. 8 B can generate from the long pulse of 275ns to 20 μ s effectively with low cost, yet it may be not suitable for producing the OTDR pulse less than 275ns.

The tunable pulsed laser source 12 of Fig. 8 B comprises SOA 202, TBF 204 and beam-splitting coupler 207, and they connect with ring topology by PMF, to form a fiber optic loop laser cavity.Coupling mechanism 207 extracts the part of light from this chamber, and typically 25 to 50%, as output.Control module 30 is coupled to SOA 202 and TBF 204 by

line

220 and 222 respectively, and it can also supply and control the bias current on the SOA 202, and regulates the wavelength of TBF 104.Control module 30 is controlled SOA 202 by line 220, opens and closes its bias current so that it generates light pulse.

Also for example, people such as Rossaro (J.Select.Topics Quantum Electronics, Vol.7, pp 475-483 (2001)), especially in its Fig. 3, suitable tunable light-pulse generator is disclosed---wherein utilize acousto-optic modulator pump light from the continuous wave tunable laser instrument.

Fig. 8 C schematically shows another suitable substituting tunable light-pulse generator, it comprises can be in wide range tuning controlled continuous wave (CW) light source 212 of live width "; and combine an independently SOA 230 ", it only serves as the amplification modulator.This CW light source comprises: broadband semiconductor optical gain medium 232 ", optical semiconductor optical amplifier (SOA) typically; And tunable bandpass filters (TBF) 234 ", its Be Controlled unit 30 (Fig. 2) control.Can be greater than the little optical signal gain of the minimum of 3-5dB near 200nm (for example, from 1250 to 1440nm, or 1440 to 1640nm).Need this minimum small-signal gain to come the compensated cavity loss, thereby form laser generation.

Continuously adjustable TBF has 30 to 80pm bandwidth (FWHM) typically, based on the bandpass filter of grating, it is used to tuning laser wavelength exactly, and light (photon) is limited in this little TBF bandwidth, to provide the optical maser wavelength with narrow linewidth exactly." " other members " of definition will comprise an output coupler (25/75 coupling mechanism typically, the 25%th, output port by reference number 136 in Fig. 8 C; But also can be 50/50 coupling mechanism, to obtain more multi output power) and an optical isolator (can be integrated in the optical gain medium, such as in the input end of SOA).

If use the PMF chamber, then without any need for other members.If but this chamber is based on SMF-28 optical fiber for example, then still needs one or two Polarization Controller, to regulate the polarization state (SOP) in this laser cavity.

SOA 230, and " produced several advantages as external modulator: one is, about 50 to 60dB high delustring (ON/OFF) rate; Second is, makes input light be amplified to 10 to 20dBm (noticing that output power intensity depends on operative wavelength) from low relatively power input (0 to 6dBm).Equipment that also it should be noted that Fig. 8 C will not produce the very narrow laser of live width.This laser linewidth depends on the TBF bandpass width consumingly.Typically, the tunable light-pulse generator of Fig. 6 can be designed as (for example has retrievable range of wavelengths near 200nm, from 1250 to 1440nm or 1440 to 1640nm)---by selecting suitable SOA (such as being the SOA at center with 1350nm and 1530nm, has the 3dB gain respectively, bandwidth is extended greater than 70nm, and maximum gain is greater than 22dB).

The equipment that also it should be noted that Fig. 8 C will not produce the very narrow laser of live width.This laser linewidth depends on the TBF bandpass width strongly.Typically, laser linewidth is about 4 to 15GHz (is 30 to 80pm for the TBF bandwidth).Yet wide laser linewidth (bandwidth) all is favourable for the application (comprising POTDR) of any OTDR, because it can reduce the coherent noise on the OTDR track.

The spectral line width of the tunable pulsed laser source in the above-mentioned various embodiment can from less than 1GHz to greater than 15GHz.In practice, the lower end of this spectral line width minimizes this by the coherent noise that makes rayleigh backscattering usually need be determined, the upper end of this live width is determined by the ability of measuring the high PMD value of appropriateness.At least approx it is favourable knowing this live width, so that use previously described live width modifying factor.At least can to regulate with known controlled way in certain is interval also be very advantageously to this laser linewidth in addition, so that overcome or significantly alleviate above-mentioned restriction about maximum detection amount PMD.The ability of the adjusting laser linewidth of if so is available, can select bigger live width so---under the situation that will measure little PMD value; And select less live width---under the situation that will measure big PMD value.Preferably, laser linewidth always is set to be approximately equal to half of selected frequency step δ v.

Those skilled in the art will recognize other alternative enforcements of these tunable optical sources.

Disturb partially

Be stressed that use the term " pseudorandom is disturbed partially " here is not need or suppose between a SOP and the next SOP to have conclusive relation.Yet this is not to say, the SOP controller 24 of physics must be veritably at random.Also can follow, for example, these SOP define uniform some lattice on Poincare sphere, have angle same between Stokes vector (Stokes vector).

Evenly distribute

" pseudorandom " SOP means, (s3) in each all is the stochastic variable that is evenly distributed between-1 and 1 to three components of the Stokes vector of the SOP on the expression Poincare sphere for s1, s2, and any in these three components all with other two uncorrelated, the mean value of product equals 0.However, be provided with no matter these SOP are on the uniform some lattice or form at random, the point on this ball must evenly distribute.

Yet if use some lattice to replace being provided with at random, this calculating or processing must not supposed between a SOP and another SOP and have conclusive relation so.Otherwise,,, will lose by this decisive relation between the track of conclusive some lattice acquisition so as may occurring in the true telecommunication link if FUT 16 moves.

The advantage of embodiment of the present invention

(1) both-end PMD measures

A. disturb the stability requirement that inclined to one side approach has loosened FUT 18 by pseudorandom, have conclusive relation because needn't suppose between the power that obtains with different SOP and/or wavelength.The FUT stability requirement that this loosens can be as small as 10ms even littler, and this depends on specific embodiments.Measurement result all is reliable for the optical fiber of any kind;

B. allow to measure the DGD at a setted wavelength place, and, when repeating, allow definite DGD that becomes with wavelength at the different wave length place, further obtain average DGD or rms DGD then;

C. particular is estimated the DGD at a setted wavelength place in the very short time (about 1ms);

If d. use narrow relatively (for example 1 to 2GHz or the narrower) tunable coherent light source of live width, the very high DGD of permission measurement FUT or overall PMD value are (for example, about 50 to 100ps), simultaneously also can be because of disturbing at random partially with the little PMD of high accuracy measurement (for example less than 0.1ps);

E. for from less than tens of minutes to several minutes overall acquisition time, can be between dynamic area from 30dB to surpassing 60dB.Can also use the estimation that obtains the DGD value less than the Measuring Time of 1 second or centisecond;

F. allow to measure the FUT that comprises the optical amplifier---as erbium doped optic fibre amplifier (EDFA) or Raman light line amplifier---in online, and, even under the situation of existence, also can obtain reliable measurement from the significant ASE light of this optical amplifier; And

G. can be designed between the FUT two ends, not have the measurement of two-way communication.

(2) single-ended overall PMD measures

A. disturb inclined to one side approach by pseudorandom and loosened FUT 18 stability requirements, between the power that obtains under different SOP and/or the wavelength, have conclusive relation because needn't suppose.This method can---for example 0.2 to 0.4s---be loosened the requirement to FUT stability in the very short period.Measurement result all is reliable for the optical fiber of any kind;

B. allow all measuring equipments all only to be positioned at the end of FUT;

C. allow to use very long pulse, for example about 1 to 20 μ s or longer---as long as OTDR can distinguish in the reflection of the local of far-end and other reflections, thereby make between the very high dynamic area of acquisition, short overall acquisition time and reduce and interfere or coherent noise.For example, for from less than 2 minutes to the overall acquisition time that surpasses 5 minutes, can obtain between very high dynamic area, from 25dB to surpassing 35dB;

If d. use the tunable pulsed laser device of narrow relatively live width (for example 1 to 2GHz or narrower), permission (is for example measured very high overall PMD value from this FUT, but still can be about 50ps or higher), because of disturbing at random partially with the little PMD of high accuracy measurement (for example less than 0.1ps); And

E. different with the situation that can use the CW laser instrument, the OTDR technology in this single-ended overall PMD mensuration of being used in can be distinguished rayleigh backscattering and in the local reflection of optical fiber far-end, so that need no longer to consider that rayleigh backscattering or other reflections such as the reflection from the connector between the fiber fragment, promptly can provide PMD measurement result very reliably.

Also should note, single-ended PMD mensuration disclosed herein can be measured from testing tool to along the PMD of any local reflection of optical fiber (for example any connector or the contact maker from the FUT), is enough to and can be suitably measured if its back reflective luminous power can be up to.

(3) single-ended accumulation PMD measures

A. disturb inclined to one side approach by pseudorandom and loosened FUT 18 stability requirements, between the track that obtains under different SOP and/or the wavelength, have conclusive relation because needn't suppose.In addition, no matter in fact disturb (method for optimizing) execution partially by I/O-SOP, still the PSP (main polarization state) by only relying on FUT " nature " disturbs partially under the situation of " ideal " FUT (as discussed previously)---and it has all obtained to loosen this advantage of FUT 18 stability requirements according to variation of wavelength and fiber lengths at random and generation equably;

B. allow to use long pulse, different with other POTDR of second type, cause:

Between the dynamic area that (i) enlarges markedly, for example, for 100 or the typical pulse length of 200ns under from being less than 10 minutes to the overall acquisition time that surpasses 30 minutes, between dynamic area from 10dB to surpassing 20dB;

(ii) reduced the OTDR coherent noise that is superimposed upon on these tracks;

(iii), increased maximum detection amount PMD for given laser spectrum live width;

C. they directly measure accumulation PMD, and these are different with first kind previously known POTDR described herein, so do not need the specific birefringence mould supposed, and especially, they are particularly suited for measuring the accumulation PMD of spin fiber;

D. produce real quantitative results; And

E. be to disturb the result of inclined to one side approach at random from measurement result of the present invention, it causes simple equation (42) significantly, and this equation is all effective for any FUT 18 and any pulse length in theory, with effective to related signal Processing.Embodiment of the present invention can be measured from the hundreds of psec and extend to above PMD value on the interval of 50 psecs, have remarkable spatial resolution and can locate high PMD fiber fragment.

With the getting in touch and distinguishing of total patented claim

The application number that on September 29th, 2006 submitted to is that the total international monopoly of PCT/CA2006/001610---content of this document is included this instructions by reference in---discloses a kind of method and facility, it carries out single-ended accumulation PMD by means of the instrument based on OTDR to FUT and measures---and right by sending a lot of group pulse series, wherein the spike train of every centering has closely alternate wavelength; And handle corresponding OTDR track, to obtain at PMD along any distance z place of this optical fiber.

Embody both-end PMD measuring method of the present invention and facility and promoted double-end measurement, wherein need to measure optical link the overall PMD and/or the DGD of one or more certain wave strong points in---it can comprise (unidirectional) optical amplifier---.In view of the above, in embodiments of the invention:

A) this measurement is measured less than reflection " leading directly to ", and pulse length is very long, makes signal to noise ratio (S/N ratio) outstanding;

B) measuring under specific wavelength should (" leading directly to " or forward) DGD, and other application then are not like this;

C) this measurement is unidirectional, therefore, if unidirectional element such as optical amplifier (comprising optical isolator) is placed in this link, then can use this method;

D) the present invention allows to measure under the following situation: when existing by the significant ASE that optical amplifier generated that gets involved;

E) the present invention allows collaborative PMD of determining and DGD (λ);

F) the invention enables the collaborative PMD that determines according to rms and average mode definition, suppose and need not to FUT performance;

G) the present invention can be adapted to be permission and monitors rapidly in the DWDM channel, to detect the flip-flop of DGD, allows thus with on possible observed system loss is related.

Embody single-ended overall PMD measurement of the present invention and solved the situation that only needs to measure overall PMD---by inserting the end of FUT.In view of the above, in such embodiment of the present invention:

A) this FUT has a local reflection with remarkable reflectivity at its far-end, and this is not the situation of the total application of quoting as proof above substantially;

B) use two detecting devices, be used for pin-point accuracy and measure reliably, it is not the situation of the total application of quoting as proof above---wherein only used single detector;

C) use long light pulse at single detector design, be used to obtain long measuring distance or dynamically high, this is not the situation of the total application of quoting as proof above---wherein only applied less than about 5 to 10 times of short optical pulse length of clapping length;

D) the back reflective pulse that is detected (" response impulse ") has and the very approaching identical duration of the pulse that is admitted to this FUT, this is different from, and the total application of quoting as proof above---wherein the back reflective signal is, corresponding to the distributed back reflective that is caused by rayleigh backscattering and along the shock response of the possible false local reflection of this FUT length.

Single-ended accumulation PMD of the present invention measurement has solved another situation of wherein using two detecting devices or two different input and output Polarization Controllers.Correspondingly, in embodiments of the invention:

A) use two detecting devices, be used for pin-point accuracy and measure reliably, this is not the situation of the total application of quoting as proof above---wherein only used single detector;

B) measure more accurately, drawing standardization power, thereby draw reliable PMD measurement result from FUT;

C), calibrate the electron device gain of two detecting devices automatically for embodiment based on PBS;

D) only use that one group of tight alternate wavelength just can obtain accumulation PMD rough, that change along with fiber lengths, yet need at least two groups in the total application of quoting as proof in the above.

Industrial usability

The full content of various patents, patent application and other documents that preamble is quoted is received by reference Enter this specification.

Although embodiment of the present invention are described and describe in detail, should be expressly understood, This only is illustrative and exemplary, and is not restrictive, and scope of the present invention is only by appended Claim limit.

Being different from known PMD measures---and known PMD measures, and its most of technology adopt two Hold mensuration, be used for current most of commercially available PMD test and measuring instruments for The field Device needs wide range of wavelengths---and both-end PMD of the present invention measures embodiment and can be employed Interval and the big range of wavelengths of small wavelength of measuring in DGD or PMD the two.

If there is any alternate channel to use, embodiment of the present invention can allow to measure and monitoring DGD or PMD in the narrow DWDM channel. Embodiment of the present invention also can allow to detect rapidly From the unexpected variation of the DGD of DWDM channel or any light path, allow thus and may observe In the system loss association.

Embodiment of the present invention allow to measure DGD or PMD in following situation: when existing by Jie During significant ASE that the optical amplifier that enters generates;

And, being different from known technology---it depends on FUT 18 in relatively long time period (allusion quotation Type ground is tens of second to a few minutes) upward keep stable---embodiment of the present invention do not need such length Stability, for example only need to be in that approximately tens of or hundreds of μ s or ms keep steady average time Fixed. This be because, obtain, corresponding to different SOP and/or wavelength (approximately tens of or Hundreds of μ s or ms are on average time) power, being taken as on the statistics is that independent (pseudorandom is disturbed Inclined to one side), and need not to suppose and have any conclusive relation between them.

And little laser linewidth of equal value can be used to obtain between high measured PMD dynamic area (for example, having about maximum detection amount PMD of 50 to 100ps that surpasses). Therefore, these are excellent The result of point is that embodying this both-end PMD measurement of the present invention can be in very short Measuring Time Inherent rise between dynamic area for this FUT measure from very little value (for example, less than 0.1ps) to the PMD of very large value (for example greater than 50 to about 100ps).

And, embody this both-end PMD of the present invention and measure and can measure with optical amplifier The PMD of FUT.

Measuring for single-ended overall PMD, be different from known PMD and measure---known PMD measures Most of technology depend on the double-end measurement method, are used for current most of commercially available PMD test and survey Measuring appratus---single-ended overall PMD of the present invention measures embodiment only needs to insert an end, that is, It is single-ended overall or total PMD measurement scheme.

And, being different from known technology---known technology relies on FUT 18 in the relatively long time Section (typically a few minutes are to dozens of minutes) is upper to keep stable---single-ended overall PMD of the present invention Measure embodiment and do not need so long stability. This be because, obtain, corresponding to not Power with SOP and/or wavelength (at about hundreds of ms on average time) is taken as statistics On be independent (pseudorandom is disturbed inclined to one side), and need not to suppose exist between them any conclusive Relation.

Use very long pulse to allow much bigger SNR, and this OTDR technology is (with CW Laser is compared) to have eliminated not be that any other light from test position (for example optical fiber end) is anti-Penetrate. And little laser linewidth of equal value can be used to obtain between high measured PMD dynamic area (for example have and approximately surpass 50 to 100ps maximum detection amount PMD). Therefore, these advantages The result be, use OTDR and long pulse, embody single-ended PMD of the present invention and measure and can close The inherent rise of the Measuring Time of the weak point of reason is measured from very little value for this FUT between dynamic area (for example, less than 0.1ps) is to the PMD of very large value (for example greater than 50 to about 100ps).

Measure for single-ended accumulation PMD, be different from known technology---known technology uses short pulse And/or depend on FUT 18 on the relatively long time period (typically a few minutes are to dozens of minutes) Keeping stablizing---single-ended accumulation PMD of the present invention measures embodiment does not need so long stablizing The property. This be because, corresponding to the OTDR of different SOP and/or wavelength (a few average time in second) Track, being taken as on the statistics is independent (pseudorandom is disturbed inclined to one side), and need not to suppose them it Between have any conclusive relation.

For given average time, use relatively long pulse to allow ratio originally can realize The SNR that SNR is much bigger. This be because: (i) luminous energy of back reflective light and pulse length are proportional; And (ii) detector electricity bandwidth can be littler, thereby allow to reduce noise. Therefore, than long pulse Length to the effect of SNR be triple with become multiplication.

With long light pulse, maximum detection amount PMD value also can be bigger because of following immediate cause: Use short pulse, " coherent noise " that is superimposed upon on the OTDR track is bigger. In order to use short arteries and veins Reduce coherent noise when rushing, " standard " scheme be increase equivalent laser linewidth (the laser natural line width, Or use and shake or other equivalent manners). This has limited maximum detection amount PMD. Therefore, use The result of the different advantages of these of long pulse is, embodies POTDR of the present invention, can be reasonably Measure that bigger accumulation PMD value---it typically appears at the farther PMD of z place in the Measuring Time Value.

In all OTDR use, carried on the back along with leaving local by the power of the light of FUT 18 back reflectives The increase of the distance that goes out to scattering and reducing is because any FUT 18 has the non-zero loss (typically at λ=1550nm place be 0.2 to 0.25dB/km). Can between the dynamic area of OTDR Be defined as, still can well measure with interior acquisition in a uncertainty that is reasonably caused by noise Maximum loss. Less than 10 to 15% situation, work as use for the uncertainty that is caused by noise The 100ns pulse of single track and 1s are during average time, and the initial testing result demonstrates about 15dB Dynamic area between. Test according to the prototype of Fig. 3 A demonstrates, and uses typical fiber loss (0.2 to 0.25dB/Km) embodies POTDR of the present invention 200ns pulse and 2s mean time Between can enough reach 70km. The expection, from the embodiment of Fig. 3,3B and 3C, can obtain close or Higher performance.

Above-mentioned advantage, the stability requirement of namely significantly loosening, the reality in response to using long pulse length The maximum detection amount of existing much bigger SNR (therefore much bigger surveying range) and reality The combination of the advantages such as PMD (such as 30 to 40ps) is so that it is especially suitable to embody POTDR of the present invention In long, mounted optical fiber, even might be " in-site measurement " that comprises those aerial optical fiber.

List of references

[1]C.D.Poole，D.L.Favin，‘Polarization-mode?dispersion?measurements?based?ontransmission?spectra?through?a?polarizer’，Journal?of?Lightwave?Technology，Vol.12(6)，pp.917-929(1994).

[2]‘Method?and?apparatus?for?measuring?polarization?mode?dispersion’，United?StatesPatent?7,227,645?B2，Jun.5，2007.

[3]N.Cyr，‘Polarization-mode?dispersion?measurement：generalization?of?theinterferometric?method?to?any?coupling?regime’，Journal?of?Lightwave?Technology，Vol.22(3)，pp.794-805(2004).

[4]‘Fiber?optical?dispersion?method?and?apparatus’，United?States?Patent?4,750,833，Jun.14，1988.

[5]P.A.Williams，A.J.Barlow，C.Mackechnie，J.B.Schlager，‘Narrowbandmeasurements?of?polarization-mode?dispersion?using?the?modulation?phase?shifttechnique’，Proceedings?SOFM，Boulder?CO，1998，pp.23-26.

[6]R.Noe，et?al，’Polarization?mode?dispersion?detected?by?arrival?time?measurement?ofpolarization-scrambled?light’，Journal?of?Lightwave?Technology，Vol.20(2)，pp.229-235(2002).

[7]‘Monitoring?Mechanisms?for?Optical?Systems’，United?States?Patent?ApplicationsPublication?US?2005/0201751?A1，Sep.15，2005.

[8]S.X.Wang，A.M.Weiner，S.H.Foo，D.Bownass，M.Moyer，M.O′Sullivan，M.Birk，M.Boroditsky，‘PMD?Tolerance?Testing?of?a?Commercial?Communication?System?Usinga?Spectral?Polarimeter’，Journal?of?Lightwave?Technology，Vol.24(11)，pp.4120-4126(2006).

[9]‘Method?and?apparatus?for?polarization?mode?dispersion?monitoring?in?a?multiplewavelength?optical?system’，United?States?Patent?7,203,428，April?10，2007.

[10]S.Wielandy，M.Fishteyn，B.Zhu，‘Optical?performance?monitoring?using?nonlineardetection’，Journal?of?Lightwave?Technology，Vol.22(3)，pp.784-793(2004).

[11]N.Kikuchi，‘Analysis?of?signal?degree?of?polarization?degradation?used?as?controlsignal?for?optical?polarization?mode?dispersion?compensation’，Journal?of?LightwaveTechnology，Vol.19(4)，pp.480-486(2001).

[12]F.Corsi，A.Galtarossa，L.Palmieri，M.Schiano，T.Tambosso，“Continuous-WaveBackreflection?Measurement?of?Polarization?Mode?Dispersion?Characterization”，IEEEPhotonics?Technology?Letters，Vol.11?No.4，April?1999，pp.451-453.

[13]A.Galtarossa，L.Palmieri，M.Schiano，T.Tambosso，“Single-End?PolarizationMode?Dispersion?Measurement?Using?Backreflected?Spectra?Through?a?LinearPolarizer”，IEEE/OSA?J.Lightwave?Technology，Vol.17?No.10，October?1999，pp.1835-1842.

[14]H.Sunnerud，B.-E.Olsson，M.Karlsson，P.A.Andrekson，J.Brentel“Polarization-Mode?Dispersion?Measurements?Along?Installed?Optical?Fibers?Using?GatedBackscattered?Light?and?a?Polarimeter”，IEEE/OSA?J.Lightwave?Technology，Vol.18No.7，July?2000，pp.897-904.

[15]H.Sunnerud，B.-E.Olsson，M.Karlsson，P.A.Andrekson，“Measurement?ofPolarization?Mode?Dispersion?Accumulation?Along?Installed?Optical?Fibers”，IEEEPhotonics?Technology?Letters，Vol.11?No.7，July?1999，pp.860-862.

[16]US?patent?No.6,229,599(A.Galtarossa).

[17]H.Dong，P.Shum，J.G.Zhou，Y.D.Gong，“Single-end?Spectral?ResolvedMeasurement?of?Polarization?Mode?Dispersion?in?Optical?Fibers”，Paper?JThA20，OpticalFiber?Communications?Conference，March?25-29，2007，Anaheim，CA，USA.

[18]US?patent?No.6,724,469(M.Leblanc).

[19]F.Corsi，A.Galtarossa，L.Palmieri，″Beat?Length?Characterization?Based?onBackscattering?Analysis?in?Randomly?Perturbed?Single-Mode?Fibers，″Journal?ofLightwave?Technology，Vol.17，No.7，July?1999.

[20]A.Galtarossa.L.Palmieri，A.Pizzinat，M.Schiano，T.Tambosso，″Measurement?ofLocal?Beat?Length?and?Differential?Group?Delay?in?Installed?Single-Mode?Fibers″，Journal?of?Lightwave?Technology，Vol.18，No.10，October?2000.

[21]A.Galtarossa，L.Palmieri，M.Schiano，T.Tambosso，″Measurement?of?Beat?Lengthand?Perturbation?Length?in?Long?Single-Mode?Fibers，″Optics?Letters，Vol.25，No.6，March?15，2000.

[22]B.Huttner，B.Gisin，N.Gisin，“Distributed?PMD?measurement?with?a?polarization-OTDR?in?optical?fibers”，Journal?of?Lightwave?Technology，Vol.17，pp.1843-1948，Oct.1999.

[23]US?patent?number?6,946,646(Chen?et?al.)

[24]US?published?patent?application?number?2004/0046955，Fayolle?et?al.

Claims

1. method of measuring at least one polarization dependent behavior of light path (FUT), this method is used near-end or near the optical input device of near-end that is connected to this light path, near and the near-end or the far-end that are connected to this light path, or the light output arrangement near-end or the far-end; This optical input device comprises the light supply apparatus of the light that is used to be supplied to the small part polarization, and is used to control the polarized state of light (I-SOP) of described partial polarization at least and described light is sent into the device of this FUT; The light output device comprises the device that is used for extracting from this FUT corresponding light, be used to analyze the analytical equipment of the light that is extracted, and pick-up unit, described pick-up unit is used to detect the analysis light (A-SOP) corresponding at least one transmission axle of this analyser device, with the coherent light power of the transmission under each wavelength in each group that at least two group wavelength are provided, wherein each organizes the low wavelength (λ in wavelength _L) and high wavelength (λ _U) closely alternate, and wherein following three conditions are not all to be satisfied simultaneously:

H. source and pick-up unit are in the same end of FUT;

I. in analysis and pick-up unit, only use a detecting device;

J. the light from light source mainly comprises transient pulse, and the space stretching of this pulse surpasses FUT and claps long 10 times;

And it is right to comprise a wavelength in wherein said group, described wavelength in each group is to all corresponding little optical frequency is poor, and at described wavelength to defining a mid point wavelength, and wherein I-SOP and A-SOP are for each the described wavelength substantially constant in each described group, and wherein between each described group, among mid point wavelength, I-SOP and the A-SOP at least one is different, and this method may further comprise the steps:

Xvi. calculate at least one poor corresponding in the power parameter that records of each wavelength of described wavelength centering for each batch total in described at least two groups, described power parameter that records and described power analyzed and detected light subsequently are proportional, define the set of at least two power parameter differences that record thus;

Xvii. calculate the mean square value of the set of described difference; And

Xviii. calculate at least one polarization-relevant FUT characteristic as at least one predefined function of described mean square value, described predefined function depends on that the little optical frequency between the described wavelength is poor, and described wavelength is corresponding to each the right wavelength in described at least two pairs of tight alternate wavelength.

2. according to the process of claim 1 wherein that described light output arrangement is connected near the far-end or far-end of this light path.

3. according to the method for claim 2, wherein:

A. each described group to comprise wavelength right, this wavelength is to having the mid point wavelength of described regulation substantially, and

B. described at least one polarization-relevant FUT characteristic is the Differential Group Delay (DGD) at described mid point wavelength place.

4. according to the method for claim 3 or 62, the wherein said power parameter that records be the standardization power T that calculates (v), and, described predefined function for little optical frequency poor (δ v) can express according to following differential formulas:

DGD (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{{< ΔT {(v)}^{2} >}_{SOP}}

Constant wherein

V is the optical frequency corresponding to described mid point wavelength.

5. according to the method for claim 3 or 62, the wherein said power parameter that records be calculate standardization power T (v), the mean-square calculation step (ii) further comprises the relative variance (σ that calculates this standardization power according to following formula _r ²(v)):

σ_{r}^{2} (v) = {(\frac{1}{σ_{20}})}^{2} [{< T {(v)}^{2} >}_{SOP} - {< T (v) >}_{SOP}^{2}]

Reference variable wherein

Then, for little optical frequency difference δ v, described predefined function is determined according to following differential formulas:

DGD (v) = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< ΔT {(v)}^{2} >}_{SOP}}{σ_{r}^{2} (v)}}

Constant wherein

V is the optical frequency corresponding to described mid point wavelength.

6. according to the method for claim 3 or 63, the wherein said power parameter that records be calculate relative power P _R(v), and the mean-square calculation step (ii) comprise following steps:

A. calculate the relative variance (σ of relative power (transmission signals relatively) _R ²(v)); And

B. calculate the ratio of described mean square deviation on described relative variance from relative power;

C. described DGD is by the function calculation of described ratio, and described function is the described predefined function that can be expressed as a differential formulas for little optical frequency difference δ v.

7. according to the method for claim 3, wherein said light supply apparatus sends coherent light at two tight alternate wavelength places, and these two wavelength are spaced apart by described little optical frequency difference, around the mid point wavelength of a regulation.

8. according to the method for claim 3, wherein

A) described optical input device sends the broadband light of polarization, and it is that the right described little optical frequency of the wavelength at center is poor that the spectral width of described broadband light has been included corresponding to the mid point wavelength with described regulation;

B) described low spaced apart by described little optical frequency difference with high wavelength, around the mid point wavelength of regulation;

C) described analysis and pick-up unit comprise the spectral filtering apparatus, and it comprises narrow band optical filter, and this filter width is poor much smaller than described little optical frequency, make therefrom select only relevant thus;

D) described spectral filtering apparatus can be operated, so that can select and detect subsequently corresponding to comprising described group right each wavelength of described wavelength.

9. method according to Claim 8, wherein said spectral filtering apparatus can be operated, allowing to select simultaneously and to detect subsequently corresponding to comprising described group right each wavelength of described wavelength, selected light through filtering is corresponding to two or more wavelength that detected by two or more detecting devices respectively subsequently.

10. according to the method for claim 3 or 5, wherein:

At each of at least two in the range of wavelengths that falls into regulation described mid point wavelength, repeat the step (i) of described method, (ii) and (iii), provide thus at least two calculate the set of DGD value, each is worth corresponding to the corresponding wavelength in these two the described mid point wavelength at least; Described optical frequency difference between the wavelength in every group needn't be identical when described method is applied in different described mid point wavelength,

Described at least one predefined function be included in the rms DGD value of DGD value at different wave length place and the average DGD value one of at least.

11. according to the method for claim 10, interior described at least two the described mid point wavelength of range of wavelengths that wherein fall into regulation comprise a large amount of mid point wavelength, these mid point wavelength are approximate to be evenly distributed on the range of wavelengths of described regulation.

12. according to the method for claim 2, wherein:

A) each group of the closely alternate wavelength of described at least two groups is limited by a mid point wavelength respectively, and at least two groups in described at least two groups have different mid point wavelength;

B) the FUT characteristic that described at least one polarization is relevant is the rmsDGD (being PMD) on the range of wavelengths of stipulating.

13. according to the method for claim 12, the wherein said power parameter that records be calculate standardization power T, wherein, for little optical frequency difference δ v, described predefined function can be expressed as following formula:

PMD = \frac{α_{ds}}{πδv} \cdot \sqrt{{< Δ T^{2} >}_{SOP; λ}}

Constant wherein

α_{ds} = \sqrt{\frac{9}{2}} .

14. according to the method for claim 12 or 62, the wherein said power parameter that records be calculate standardization power T (v), and wherein the mean-square calculation step (ii) comprises the relative variance (σ that calculates this standardization power according to following formula _r ²):

σ_{r}^{2} = {(\frac{1}{σ_{20}})}^{2} [{< T^{2} >}_{SOP; λ} - {< T >}_{SOP; λ}^{2}]

Reference variable wherein

σ_{20}^{2} = 1 / 12,

PMD = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< Δ T^{2} >}_{SOP; λ}}{σ_{r}^{2}}}

Constant wherein

α_{ds} = \sqrt{\frac{9}{2}} .

15. according to the method for claim 14 or 63, the wherein said power parameter that records be calculate relative power P _R, and the mean-square calculation step (ii) comprises following steps:

A) relative variance (σ of the relative transmission signals of calculating _R ²(v)); And

B) calculate the ratio of described mean square deviation to described relative variance, described rms DGD is calculated as the function of described ratio, predefined function as described, and described predefined function is determined according to following differential formulas for little optical frequency difference δ v:

PMD = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< Δ P_{R}^{2} >}_{SOP; λ}}{σ_{R}^{2}}}

Relative variance (the σ of standardization power wherein _R ²) be defined as:

σ_{R}^{2} = {(\frac{1}{σ_{20}})}^{2} [{< P_{R}^{2} >}_{SOP; λ} - {< P_{R} >}_{SOP; λ}^{2}] .

16. according to the method for claim 2, wherein:

In each step of at least one spectrum obtaining step, detected and the storage of at least one quasi-continuous transmission coherent light power that changes along with optical frequency, be used for described step (i) and further analyze, described optical frequency is crossed over the range of wavelengths of a regulation

A) from described transmission coherent light power, calculate the described power parameter that records;

B) I-SOP and A-SOP not, one or both of because of optical frequency different, and if this each variation---existence---be slowly so that for each described closely alternate set of wavelengths, I-SOP and A-SOP are basic identical respectively.

17. according to the method for claim 2, wherein

A) to obtain be that a spectrum obtains to described at least one spectrum; And

B) the relevant FUT characteristic of described at least one polarization is the rms DGD (being PMD) on the range of wavelengths of described regulation.

18. according to the method for claim 16 or 62, wherein the power parameter that records of step b) is a standardization power, it is with that analyzed and luminous power that detected subsequently is proportional.

19. according to the method for claim 18 or 63, wherein the power parameter that records of step b) is a relative power, it is with that analyzed and luminous power that detected subsequently is proportional.

20. method according to claim 18, it is that at least two spectrum obtain that wherein said at least one spectrum obtains, the I-SOP of some optical frequency of storing at least in wherein obtaining corresponding at least one spectrum and A-SOP one or both of are different from the I-SOP of the respective optical frequency of being stored and A-SOP one or both of respectively at least one scans for the second time, described at least one predefined function comprise following at least one

A. the regulation range of wavelengths on rms DGD value; And

B. when described some optical frequency of storing at least during corresponding to described mid point wavelength, the DGD at least one described mid point wavelength place.

21. according to the method for claim 20, wherein

A) described at least one the described mid point wavelength b) is a mid point wavelength; And

B) described spectrum obtain included corresponding to the right described little optical frequency of the wavelength that is the center with described mid point wavelength poor.

22. according to the method for claim 16, wherein

A) described not, one or both of, be not; And

23. according to the method for claim 16, wherein said light supply apparatus can be operated, so that scan the optical frequency of described coherent light substantially continuously, described scanning makes it possible to carry out described spectrum and obtains, and included the range of wavelengths of regulation.

24. according to the method for claim 16, wherein

A) described light supply apparatus sends the broadband light of polarization, and the spectral width of described broadband light has been included the spectrum range of regulation;

B) described analysis and pick-up unit comprise the spectral filtering apparatus, and it comprises narrow band optical filter, and this filter width is poor much smaller than described little optical frequency so that therefrom select only relevant; And

C) described spectral filtering apparatus is to operate, and scanning (scanning) substantially continuously, selecting in succession and to detect subsequently corresponding to comprising described group right each wavelength of described wavelength, described scanning makes it possible to carry out described spectrum and obtain.

25. according to the method for claim 16, wherein

A) described light supply apparatus sends the broadband light of polarization, and the spectral width of described broadband light has been included the spectrum range of regulation; And

B) described spectral filtering apparatus, comprise spectrometer, the spectrum of spatially separating incident light detects corresponding to the right wavelength of described wavelength in described two groups simultaneously to allow pick-up unit at least, and different wavelength is optionally detected by two or more detecting devices respectively.

26. according to the process of claim 1 wherein that described light extraction and treating apparatus are connected near the light path this FUT near-end or the near-end, and this FUT far-end or near the local reflection is provided.

27. according to the method for claim 26, wherein:

A) each group in the closely alternate wavelength of described at least two groups is limited by separately mid point wavelength, and at least two groups in described at least two groups have different mid point wavelength; And

B) the relevant FUT characteristic of described at least one polarization is the rms DGD (being PMD) forward on the range of wavelengths of regulation.

28. method according to claim 26, wherein each group comprises a series of light pulse, each light pulse in each series has less than the space of this FUT length extends, and has essentially identical wavelength with other pulses that are arranged in same described each series, and wherein, some light pulse at least from each series average determined the described power parameter that records.

29. according to the method for claim 27 or 62, the wherein said power parameter that records be calculate standardization power T, and for little optical frequency difference δ v, described predefined function is determined according to following differential formulas:

PMD = \frac{α_{rt} \cdot α_{ds}}{πδv} \cdot \sqrt{{< Δ T^{2} >}_{SOP; λ}}

Wherein come and go the factor

Constant α _DsDepend on detected from the light of propagating forward of light source or from the back reflective light (coming and going light) of light source the light path of process.

30. method according to claim 27 or 62, the wherein said power that records be calculate standardization power T, and the unpolarized noise that may exist in the signal that (ii) detected by following steps compensation of mean-square calculation step---such as spontaneous radiation (SE) light:

A) relative variance (σ of normalized transmission signals _r ²); And

B) calculate the ratio of described mean square deviation to described relative variance, described rms DGD is calculated as the function of described ratio, and this function is predefined function as described, and it is determined according to following differential formulas for little optical frequency difference δ v:

PMD = \frac{α_{rt} \cdot α_{ds}}{πδv} \cdot \sqrt{\frac{{< Δ T^{2} >}_{SOP; λ}}{σ_{r}^{2}}}

Wherein

The relative variance of standardization power is defined as,

σ_{r}^{2} = {(\frac{1}{σ_{10}})}^{2} [{< T^{2} >}_{SOP; λ} - {< T >}_{SOP; λ}^{2}]

Wherein

Come and go the factor

31. according to the method for claim 27 or 63, the wherein said power that records be calculate relative power P _R, and the mean-square calculation step (ii) comprises:

A) relative variance (σ of the relative transmission signals of calculating _R ²); And

B) calculate the ratio of described mean square deviation to described relative variance, described rms DGD is calculated as the function of described ratio, and this function is predefined function as described, and it can be expressed as a differential formulas for little optical frequency difference δ v.

32. according to the method for claim 26, wherein

A. each described group to comprise wavelength right, this wavelength is to having the mid point wavelength of defined substantially, and

B. the relevant FUT characteristic of described at least one polarization is the Differential Group Delay (DGD) at described mid point wavelength place.

33. according to the method for claim 32 or 62, the wherein said power parameter that records be calculate standardization luminous power T (v), and for little optical frequency difference δ v, described predefined function is according to following differential formulas:

{DGD (v)}_{RoundTrip} = \frac{α_{ds}}{πδv} \cdot \sqrt{{< Δ {T (v)}^{2} >}_{SOP}}

Constant α wherein _DsDepend on detected from the light of propagating forward of light source or from the back reflective light (coming and going light) of light source the light path of process.

34. method according to claim 32 or 62, the wherein said power parameter that records be calculate standardization luminous power T (v), and the unpolarized noise that may exist in the signal that (ii) detected by following steps compensation of mean-square calculation step---such as spontaneous emission light:

A) relative variance (σ of normalized transmission signals _r ²(v)); And

{DGD (v)}_{RoundTrip} = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< Δ T^{2} (v) >}_{SOP; λ}}{σ_{r}^{2} (v)}}

Wherein the relative variance of standardization power is defined as,

σ_{r}^{2} (v) = {(\frac{1}{σ_{10}})}^{2} [{< T^{2} (v) >}_{SOP} - {< T (v) >}_{SOP}^{2}]

Constant wherein

And constant α wherein _DsDepend on detected from the light of propagating forward of light source or from light source back reflective light (coming and going light) the light path of process.

35. according to the method for claim 32 or 63, the wherein said power parameter that records be calculate relative power P _R(v), the mean-square calculation step (ii) comprises following steps:

B) calculate the ratio of described mean square deviation to described relative variance, described DGD is calculated as the function of described ratio, and this function is predefined function as described, and it can be expressed as a differential formulas for little optical frequency difference δ v.

36. according to the method for claim 32, wherein:

A. at each wavelength at least two in the range of wavelengths that the drops on regulation described mid point wavelength, repeat described method, provide thus at least two calculating of corresponding at least two described mid point wavelength places round total DGD value; Described optical frequency difference between the wavelength in every group needn't be identical when described method is applied in different described mid point wavelength; And

B. the relevant FUT characteristic of described polarization also comprise following one of at least:

I. average on the range of wavelengths of regulation comes and goes overall DGD value, wherein this mean value be from the round overall DGD that individually records (v) calculate root mean square (rms) and mean value arbitrary or the two, and

Ii. corresponding to average (forward) of the range of wavelengths of regulation overall DGD value---wherein this on average be root mean square (rms) and average arbitrary or the two---be by the suitable round factor is included, from the round overall DGD that individually records (v, z) calculate.

37. according to the method for claim 36, interior described at least two the described mid point wavelength of range of wavelengths that wherein drop on regulation comprise a large amount of mid point wavelength, these mid point wavelength are approximate to be evenly distributed on the range of wavelengths of described regulation.

38. according to the process of claim 1 wherein:

A. described light output arrangement is connected near the near-end or near-end of this FUT light path;

B. it is right that each group comprises at least one wavelength of a series of light pulses, and each series has identical I-SOP;

C. the light pulse in each right series of wavelength has essentially identical wavelength;

D. the described power parameter that records is the back reflective power that changes along with the distance on this FUT that is detected, and the described power parameter that records is determined by following:

I. in some light pulse at least in each the serial light pulse in each described group each, analyze and detect subsequently light---it comprises at least one polarized component of the signal of back reflective as a result that is caused by Rayleigh scattering and/or discrete reflectance along this FUT---so that corresponding shock response to be provided, described at least one polarized component is identical for each the described series in described group, and each shock response is converted to the electric shock response signal of a correspondence;

Ii. for each the described light pulse series in each described group, the electric shock response signal of sampling and average described some light pulse at least is to provide the OTDR that changes along with time delay track;

The described OTDR track that iii. will change along with time delay is converted to the OTDR track of expression back reflective power that detects of variation along with distance.

39. according to the method for claim 38, wherein:

A. each group in the closely alternate wavelength of described at least two groups is limited by a centre wavelength separately respectively, if this group only comprises two series corresponding to each closely alternate wavelength, then described centre wavelength is the mid point wavelength, and at least two groups in described at least two groups have different centre wavelength, and

B. the relevant FUT characteristic of described at least one polarization is the accumulation PMD value corresponding to the distance z on this FUT on the range of wavelengths of regulation, and described accumulation PMD value is to estimate from the round DGD of accumulation rms at the range of wavelengths of identical described regulation.

40. according to the method for claim 39 or 62, the wherein said power parameter that records be calculate the standardization power T (z) that changes along the distance z of this FUT, described predefined function is to determine according to following differential formulas for little optical frequency difference δ v:

PMD (z) = \frac{α_{rt} \cdot α_{ds}}{πδv} \cdot \sqrt{{< Δ T^{2} (z) >}_{SOP; λ}}

Wherein come and go the factor And constant α wherein _DsDepend on light that the light that detected still detected from the light of propagating forward of light source from the back reflective light (coming and going light) of light source the light path of process.

41. method according to claim 39 or 62, the wherein said power parameter that records be calculate standardization power T (z), and the mean-square calculation step is (ii) extended the local depolarization to the back reflective signal that causes by following steps compensation by the space of light pulse:

A) relative variance (σ of normalized transmission signals _r ²(z)); And

B) calculate the ratio of described mean square deviation to described relative variance, described accumulation PMD is calculated as the function of described ratio, this function predefined function as described, and it can be expressed as following differential formulas for little optical frequency difference δ v:

PMD (z) = \frac{α_{rt} \cdot α_{ds}}{πδv} \cdot \sqrt{\frac{{< Δ T^{2} (z) >}_{SOP; λ}}{σ_{r}^{2} (z)}}

Constant α wherein _DsDepend on light that the light that detected still detected from the light of propagating forward of light source from the back reflective light (coming and going light) of light source the light path of process, come and go the factor

The relative variance of this standardization power is defined as,

σ_{r}^{2} (z) = {(\frac{1}{σ_{10}})}^{2} [{< T^{2} (z) >}_{SOP; λ} - {< T (z) >}_{SOP; λ}^{2}]

Constant wherein

σ_{10}^{2} = \frac{4}{45} .

42. according to the method for claim 39 or 63, the wherein said power parameter that records be calculate relative power P _R(z), and the mean-square calculation step (ii) comprise following steps:

C) relative variance (σ of the relative transmission signals of calculating _R ²(z)); And

D) calculate the ratio of the mean square deviation of relative power to described relative variance, described DGD is calculated as the function of described ratio, this function predefined function as described, and it is determined according to differential formulas for little optical frequency difference δ v.

43. according to the method for claim 38, wherein:

A) each described group wavelength that comprises the mid point wavelength that has described regulation substantially is right; And

B) the relevant FUT characteristic of described at least one polarization is the Differential Group Delay (DGD) at described mid point wavelength place.

44. method according to claim 43 or 62, the wherein said power parameter that records be calculate standardization back reflective power T (v, z), and the mean-square calculation step (ii) extend the local depolarization that causes the back reflective signal by the space of light pulse by following steps compensation:

A. relative variance (the σ of normalized transmission signals _r ²(z, v)); And

B. calculate the ratio of described mean square deviation to described relative variance, described DGD is calculated as the function of described ratio, this function predefined function as described, and it is determined according to following differential formulas for little optical frequency difference δ v:

{DGD}_{RoundTrip} (z, v) = \frac{α_{ds}}{πδv} \cdot \sqrt{\frac{{< ΔT {(z, v)}^{2} >}_{SOP}}{σ_{r}^{2} (z, v)}}

Constant α wherein _DsDepend on light that the light that detected still detected from the light of propagating forward of light source from the back reflective light (coming and going light) of light source the light path of process, the relative variance of this standardization power is defined as,

σ_{r}^{2} (z, v) = {(\frac{1}{σ_{10}})}^{2} [{< T^{2} (z, v) >}_{SOP; λ} - {< T (z, v) >}_{SOP; λ}^{2}]

Constant wherein

σ_{10}^{2} = \frac{4}{45} .

45. according to the method for claim 43 or 63, the wherein said power parameter that records be calculate relative power P _R(z v), can also comprise an additional calculation procedure in (ii) in the signal that is detected, this step comprises:

A) relative variance (σ of the relative transmission signals of calculating _R ²(z, v)); And

B) calculate the ratio of described mean square deviation to described relative variance, described DGD is calculated as the function of described ratio, this function predefined function as described, and it is represented as a differential formulas for little optical frequency difference δ v.

46. according to the method for claim 43, wherein:

A) at least two described mid point wavelength in the range of wavelengths that drops on regulation each, repeat described method, provide thus at least two calculating of corresponding at least two described mid point wavelength places round DGD value; Described optical frequency difference between the wavelength in every group needn't be identical when described method is applied in different described mid point wavelength; And the relevant FUT characteristic of described polarization also comprise following at least one:

Iii. average DGD (z) value that comes and goes on the range of wavelengths of regulation, wherein this mean value be from come and go DGD (v, z) calculate root mean square (rms) and mean value arbitrary or the two; And

Iv. corresponding to average (forward) DGD value of range of wavelengths of regulation---wherein this on average be root mean square (rms) thereby and average arbitrary or the two---be calculate by the suitable round factor is included round DGD (v z) is worth.

47. according to the method for claim 46, interior described at least two the described mid point wavelength of range of wavelengths that wherein drop on regulation comprise a large amount of mid point wavelength, these mid point wavelength are approximate to be evenly distributed on the range of wavelengths of described regulation.

48. according to claim 3,10,12,27,32, arbitrary method of 39 and 43, wherein each I-SOP and A-SOP arbitrary or the two---each is corresponding at least one described set of wavelengths---can significantly be different from its forerunner person or succession separately respectively, and on Poincare sphere randomly or pseudorandom ground select, these are different to make for described group enough big quantity K, and each I-SOP and A-SOP arbitrary or the two are evenly distributed in respectively on separately the Poincare sphere substantially.

49. according to claim 3,10,12,16,27,32, arbitrary method of 39 and 43, wherein each I-SOP and A-SOP arbitrary or the two---each is corresponding at least one described set of wavelengths---only is different from its pioneer and succession separately respectively slightly, these are different to make for described group enough big quantity K, and each I-SOP and A-SOP arbitrary or the two are evenly distributed in respectively on separately the Poincare sphere substantially.

50. according to claim 3,10,12,27,32, arbitrary method of 39 and 43, wherein each I-SOP and A-SOP arbitrary or the two comprise four distinct I-SOP and A-SOP respectively, and described four distinct I-SOP and A-SOP represent the point that is equally spaced substantially on the Poincare sphere respectively.

51. according to the method for claim 48, wherein said quantity K is greater than 10.

52. according to the method for claim 49, wherein said quantity K is greater than 500.

53. according to claim 4,5,13,14,15,29,30,33,34,40, arbitrary method of 41 and 44, wherein for little optical frequency difference δ v, described predefined function can be expressed as any such formula: the numerical result that it provides and the difference of described following differential formulas fall in the acceptable scope.

54. according to claim 3,10,13,27,32, arbitrary method of 39 or 43, wherein said at least two group wavelength comprise the described set of wavelengths of K quantity, it is right that each set of wavelengths comprises at least one wavelength, wherein k group by SOP to (I-SOP _k, A-SOP _k) sign, wherein (I-SOP _k) refer to send into the polarization state of this FUT, (A-SOP _k) refer to described at least one transmission axle of analytical equipment, so that for right corresponding to K described group K, at least two of described SOP centering are different substantially with other of described SOP centering.

55. according to the method for claim 54, wherein send into additional K described set of wavelengths, at least two of described SOP centering are most of K described SOP centering.

56. according to claim 3,10,12,27,32, arbitrary method of 39 or 43, wherein each described closely alternate set of wavelengths comprises, detect at least one additional repetition described wavelength centering each wavelength---it is right corresponding to initial first wavelength, wherein the I-SOP and the A-SOP of each of the wavelength centering of these additional repetitions are basic identical in each described group, the calculating of the relevant FUT characteristic of this at least one described polarization comprises, at the right detection signal of wavelength of these additional repetitions.

57. method according to claim 56, in pick-up unit, provide right to each the described wavelength of the wavelength centering of described additional repetition and corresponding initial wavelength, pick-up unit separates the described transmission light of analyzing, so that comprise the identical transmission light of being analyzed of two parts, described two parts (analysis light) are detected by two detecting devices respectively.

58. it is, wherein, in pick-up unit,,, one after the other detected in time with respect to each wavelength of the initial wavelength centering of described correspondence from described each wavelength of the wavelength centering of the described additional repetition of light supply apparatus according to the method for claim 57.

59. method according to claim 56, wherein, in pick-up unit, from described each wavelength of the wavelength centering of the described additional repetition of light supply apparatus, with respect to the different slightly wavelength of the initial wavelength centering of described correspondence, one after the other detected in time.

60. according to claim 3,10,12,27,32, arbitrary method of 39 or 43, means wherein are provided, make the transmitting optical power that detected with described group in the highest wavelength, lowest wavelength and middle wavelength---when comprising more than two wavelength for described group---in corresponding one relevant.

61. according to the method for claim 60, wherein said relevant apparatus comprises that its coding comes the coding that light pulse is carried out by using the arbitrary of amplitude or pulsed frequency or the two.

62. according to claim 4,5,13,14,15,29,30,33,34,40, arbitrary method of 41 or 42, wherein the power parameter that records of step (i) is standardization power T, it is determined by one of following method with that analyze and luminous power that detected subsequently is proportional:

A) use a detecting device to detect a polarized component of this luminous power easily, then, each wavelength at the coherent light in having each described set of wavelengths of at least two wavelength, by with average divided by at least some---preferably whole---power of the coherent light on the same group not of the power of this coherent light, obtain standardization power respectively;

B) use two detecting devices to detect two orthogonal polarization components of this luminous power simultaneously easily, then, each wavelength at coherent light, by will be corresponding in the power of this two different polarization components that detected of this coherent light at least one divided by power sum corresponding to this two different polarization components that detected of this coherent light, perhaps by will corresponding to the power of this two different polarization components that detected of this coherent light multiply by the later difference of weight coefficient divided by power sum corresponding to this two different polarization components that detected of this coherent light, obtain the standardization power of each wavelength of coherent light;

C) use two detecting devices to detect a polarized component and a luminous power that directly is directly proportional with the light of exporting from this FUT easily, then, by at first will be corresponding to from the power of the relevant light wavelength of a detected luminous power of polarized component of light power divided by the coherent light of the luminous power that directly is directly proportional corresponding to light with output, the ratio of representing the relative power of this coherent light with acquisition, and, obtain standardization power corresponding to each wavelength of coherent light with average divided by at least some---preferably whole---relative powers on the same group not of the described relative power of this coherent light;

D) use a detecting device to add an optical switch, detect two orthogonal polarization components of this light at different time by same detecting device, wherein this optical switch is used to these two orthogonal polarization components with light and is routed to identical detecting device and detects, then, by will be corresponding in the power of this two different orthogonal components that detected of this coherent light at least one divided by power sum corresponding to this two different orthogonal components that detected of this coherent light, perhaps by will corresponding to the power of this two different polarization components that detected of this coherent light multiply by the later difference of weight coefficient divided by power sum corresponding to this two different polarization components that detected of this coherent light, obtain standardization power at each wavelength of coherent light;

E) use a detecting device to add an optical switch, detect a polarized component and the optical power that direct and total light intensity is directly proportional of this light at different time by same detecting device, wherein this optical switch is used to be routed to same detecting device with a polarized component with the direct proportional luminous power of exporting from this FUT of light, then, by at first will be corresponding to from the power of the relevant light wavelength of a detected luminous power of polarized component of light power divided by the coherent light of the luminous power that directly is directly proportional corresponding to total light intensity with output, the ratio of representing the relative power of this coherent light with acquisition, and, obtain standardization power corresponding to each wavelength of coherent light with average divided by at least some---preferably whole---relative powers on the same group not of the described relative power of this coherent light.

63. according to claim 6,15,19,31,35, arbitrary method of 42 and 45, wherein the power parameter that records of step (i) is relative power P _R, it is proportional with the institute luminous power analyzing and detected subsequently, and is definite by one of following method:

A) use detecting device alone to detect a polarized component of this luminous power easily, then, each wavelength at the coherent light in having each described set of wavelengths of at least two wavelength, by with average divided by at least some---preferably whole---power of the coherent light on the same group not of the power of this coherent light, obtain standardization power;

B) use two detecting devices to detect two orthogonal polarization components of this light simultaneously easily, then, each wavelength at coherent light, by will be corresponding in the power of this two different polarization components that detected of this coherent light at least one divided by power sum corresponding to this two different polarization components that detected of this coherent light, perhaps by will corresponding to the power of this two different polarization components that detected of this coherent light multiply by the later difference of weight coefficient divided by power sum corresponding to this two different polarization components that detected of this coherent light, obtain the relative power of each wavelength of coherent light;

C) luminous power using two detecting devices to detect a polarized component and directly be directly proportional with the light of exporting from this FUT, then, by will corresponding to from the power of a detected coherent light of polarized component of light divided by corresponding to the power of the coherent light of the direct directly proportional luminous power of total light intensity of output to obtain relative power corresponding to each wavelength of this coherent light, obtain to represent the ratio of the relative power of this coherent light;

D) use a detecting device to add an optical switch, detect two orthogonal polarization components of this light at different time by same detecting device, wherein this optical switch is used to these two orthogonal polarization components of light are routed to a described detecting device, then, by will be corresponding in these two the pairing power of different orthogonal component that detected of this coherent light at least one divided by power sum corresponding to this two different orthogonal components that detected of this coherent light, perhaps by will corresponding to the power of this two different polarization components that detected of this coherent light multiply by the later difference of weight coefficient divided by power sum corresponding to this two different polarization components that detected of this coherent light, obtain relative power at each wavelength of coherent light;

E) use a detecting device to add an optical switch, detect two orthogonal polarization components of this light at different time by a described detecting device, wherein this optical switch is used to be routed to a described detecting device with a polarized component with luminous power that light from this FUT output directly is directly proportional, then, by will corresponding to from the power of each wavelength of the coherent light of a detected luminous power of polarized component of light divided by the power of the coherent light of the luminous power that directly is directly proportional corresponding to total light intensity with output to obtain relative power corresponding to each wavelength of this coherent light, obtain to represent the ratio of the relative power of this coherent light.

64. according to claim 3,10, arbitrary method of 12 or 16, wherein:

A) at least one transmission axle of analyser device comprises the transmission axle of two or more independent linearities; And

B) the coherent light power that is transmitted from a plurality of described transmission axles is substantially side by side detected by the detecting device of the correspondence in the described detector means.

65. according to claim 29,30,33,34,40, arbitrary method of 41 or 44, wherein, if from light source and send into the light of propagating forward of this FUT and the back reflective light that detected the two through a public polarization scrambler, then constant

If described light of propagating forward and the described back reflective light that detects are through polarization scrambler independently, then

66. be used to measure the surveying instrument of at least one polarization dependent behavior of light path (FUT), comprise:

This optical input device comprises:

Light supply apparatus, it is used for supplying the light of the partial polarization at least under each ripple of at least two set of wavelengths, and

The SOP control device, it is used to control the polarized state of light (I-SOP) of described partial polarization at least, and described light is sent into this FUT, wherein the low described wavelength (λ in each described set of wavelengths _L) and high described wavelength (λ _U) closely alternate,

The SOP of the light of sending into and A-SOP are for each the described wavelength substantially constant in each described group, and wherein, between each described group, at least one among mid point wavelength, I-SOP and the A-SOP is different, and

The light output device comprises:

K. source and pick-up unit are in the same end of FUT;

L. in analysis and pick-up unit, only use a detecting device;

M. the light from light source mainly comprises transient pulse, and the space of this pulse extends beyond FUT and claps long 10 times;

This treating apparatus is configured and can operates to be used for:

Xix. calculate at least one poor corresponding to the power parameter that records of each wavelength of described wavelength centering for each batch total in described at least two groups, described power parameter that records and described power analyzed and detected light subsequently are proportional, define the set of at least two power parameter differences that record thus;

Xx. calculate the mean square value of the set of described difference; And

Xxi. calculate at least one polarization-relevant FUT characteristic as at least one predefined function of described mean square value, described predefined function depends on that the little optical frequency between the described wavelength is poor, and described wavelength is corresponding to each the right wavelength in described at least two pairs of tight alternate wavelength; And

Xxii. export the value of described at least one polarization-relevant FUT characteristic, so that show, transmit or further handle.

67. be used for continuously and repeatedly produce at two or more closely alternate wavelength places the light source of coherent lights, this equipment comprises:

Optical gain medium;

At least two laser cavitys, the part of their each laser cavity is shared in each chamber, comprises described optical gain medium;

68. equipment according to claim 67, if wherein described at least one output coupler is two or more output couplers, then the light of each extraction from described two or more output couplers one after the other is attached in the common optical path, to provide single output from described light source.

69. according to the equipment of claim 67 or 68, wherein:

This multi-channel optical retention device is that optics isolating switch dish and being used to drives this and coiling the motor that passes this disk center and rotate perpendicular to the axle on the plane of this dish, and described dish comprises some holes of arranging in a predefined manner;

The light that separates on described space part spatially is distributed, and passes described dish to allow in each part one substantially without hindrance along with the variation of the angle of rotation of this dish;

Each light part is passed this dish substantially without hindrance, finishes passing through of its each chamber then; And

This dish is with the rotation of the speed of approximately constant, partly pass corresponding to the light of each described laser cavity one after the other and repeatedly allowing, and rule is provided with the time interval that repeats, in the described time interval, do not have light to pass described dish;

Then, this equipment can be operated one after the other and repeatedly to produce two tight alternate wavelength at least.