BR102012023106A2 - PMD LOCALIZATION AND ESTIMATE METHOD IN OPTICAL LINES - Google Patents

Abstract

METHOD OF LOCALIZATION AND ESTIMATION OF PMD IN OPTICAL FLANTS, based on the analysis of the degree of roughness of pOTDR traces obtained from the emission, by an OTDR (201), of a pulse sequence (301, 304) of various temporal widths, routed to a fiber optic link (205). The Rayleigh effect backslash signal (207, 208, 210, 211) passes an optical polarizer (209) and the roughness of its trace is compared to a predefined threshold level. Pulses are ordered in a sequence ranging from the largest time-width pulses (304) to the smallest time-width pulses (301).

Description

METHOD OF LOCALIZATION AND ESTIMATION OF PMD IN OPTICAL LINES.

Field of Application The present invention applies to the field of Telecommunications Engineering, specifically referring to a method which enables the location and estimation of PMD in various types of fiber optic links using the pOTDR technique. This method makes it possible to locate high PMD stretches in field-installed fiber optic links, without the need to disconnect sections of said links.

For a better understanding of this descriptive report, here are some acronyms, expressions and terms used in it. DGD (Differential Group Delay) - Fiber group differential delay, also known as first order PMD.

Rayleigh scattering - A phenomenon that occurs when a beam or light pulse collides with a particle whose diameter is smaller than the wavelength of light, resulting in scattering in a random direction.

Rayleigh Backscatter - Refers to the portion of light that undergoes Rayleigh scattering and is reflected back to the release point of the propagating light pulse on a fiber optic link. In the case of the object of the present invention, the release point of the pulse is a Commercial OTDR or an OTDR module. Optical Time Domain Reflectometei (OTDR) - Optical Time Domain Reflectometer. Polarization Mode Dispersion (PMD) - Polarization Mode Dispersion (PMD) refers to the differences in the propagation velocities of two polarizations along the optical fiber, caused by imperfections and asymmetries in its core. pOTDR (OTDR polarization) - Time domain optical reflector of polarized signals. PPCR - First pulse with roughness SOP (State of Polarization) - Polarization state UPSR - Last pulse without roughness VOS (Variation of Signal) - Signal Variation State of the Art The use of displaced dispersion and dispersion in compensating fibers has minimized the effect of chromatic dispersion in the bandwidth of an optical link. Polarization mode dispersion (PMD) has therefore become the most severe limiting factor in high speed optical communication systems.

There are several PMD measurement techniques, but they only allow the overall value to be measured and do not give information on the distribution of PMDs along the fiber length. Measuring the spatial distribution of PMD, however, is important. Such a measure would, in fact, make it possible to locate the bad portions on an optical link with a high PMD value. This can be essential when maintaining and updating networks. The most commonly used equipment to verify the quality of the installed or to be installed optical fiber in optical cables is the time domain optical reflectometer (OTDR). OTDR's operation is based on the analysis of the portion of light that returns to the release point of the light pulse by Rayleigh scattering. To this end, OTDR transmits pulses of a given time width to the optical fiber link and, as light propagates along the fiber, a portion of it is backscattered by the Rayleigh effect.

This backscattered light signal is attenuated by the fiber optic link and captured by the OTDR allowing the attenuation coefficient of the optical fiber to be calculated along its length. OTDR also measures point reflections generated along the fiber, and this feature is of fundamental importance for fault detection such as fiber breakage, poorly made optical splices, faulty connections, sharp bends, etc.

OTDRs have long been used for surveying the attenuation characteristics of fiber optics before and after installation, as well as for quality control purposes in fiber and optical cable factories. In telecom operators, OTDRs are valuable tools to ensure preventive maintenance of installed links.

On the other hand, OTDR does not allow analysis of signal polarization states along the fiber. This analysis is of great interest because it allows to determine, for example, high PMD stretches in optical links. Some proposed equipment intended to perform this analysis are those known as pOTDR (polarization OTDRs). However, such equipment needs dedicated detection, presents complexity of instrumental implementation or complexity in the treatment of the obtained signals, as explained below. Most of the proposed techniques for measuring distributed PMD are based on time domain optical polarization reflectometry (pOTDR). The concept of pOTDR was introduced in the 1980s (A.Rogers, “Polarization-optical time domain refiectometry: a technique for measurement of filed distributions,” Appled Optics, vol.20 / pp. 1060-1074,1981). Such a technique basically consists of measuring the signal Rayleigh backscatter properties when an optical pulse propagates in the fiber.

Since then some pOTDR techniques have been developed for PMD characterization. Most of these techniques allow a measurement of the bias stroke length distribution over an optical link rather than determining the PMD value itself.

However, a pOTDR-based experimental arrangement has been developed that allows the measurement of accumulated PMD along the fiber (see H.Sunnerud et al, Polarization Mode Dispersion measurements along installed fibers using gated back scattered light and a polarimeter, Journal of Lightwave Technol., Vol. 18, p. 897/904, 2000). However, this method requires pOTDR measurements at various wavelengths.

Another technique based on measuring polarization degree using a pOTDR configuration was presented by M. Leblanc in “Distributed detection of high PMD sections on installed fibers using an OTDR * polarization” (Proc. OFMC 01, pag.155 / 162, 2001 ) which is related to US patent application 2003/0174312. This technique does not allow the quantification of the PMD, but allows for distributed information about the PMD level. In addition, this technique requires complete measurement of the bias state of the backscattered signal.

More recently, Galtarossa described another technique for measuring PMD in which the coupling length is determined from the birefringence vector correlation (see “Measurement of birefringence correlation length in long single-mode fibers”, Galtarossa et al., Opt. Lett., Vol.26, pag.962 / 964, 2001), which also requires measurement of the bias states of the backscattered signal.

In "Polarization mode dispersion mapping in optical fibers with a Polarization OTDR", M. Wuilpart, G. Ravet, P. Megret and M. Blondel (Photonic Technology Lett., Vol.14, page 1716/1718, 2002), The authors describe an analysis of the pOTDR signal that allows PMD mapping on a fiber optic link by quantifying the PMD on each fiber of the link.The main advantage of this technique is that it does not require complete measurement of the bias state of the backscattered signal. and uses only one linear polarizer in the input fibers, which is quite simple to implement.The determination of bias beating length, coupling length, and finally PMD is based on the analysis of the statistical properties of the ends contained in the trace. The method described above has some important limitations and disadvantages, providing only approximate values for beat, coupling, and PMD lengths. pOTDR presented were assumed to be ideal.

This method assumes ideal characteristics of the various components of the experimental installation. In practice, the proposed arrangement has imperfections that cause optimal pOTDR signal distortion.

Therefore, the method as such cannot be directly applied. The pOTDR signal is also affected by detector noise and coherence residual noise, which adds a series of minimum and maximum levels for pOTDR tracking. The final measured PMD value can therefore be wrong. Publication WO 2005/041449 Al, related to that article, uses pOTDR traces obtained by emitting optical pulses and measuring the backscattered signal (pOTDR trace) after passing through a polarizer. The technique described herein allows to determine the length of the fiber optic link, the average power between two successive minima of said backscattered signal and the number of maximums per unitary fiber length. Iteratively then determines a bias beating length range, a bias mode coupling parameter range, until the length of said bias range is less than a predetermined value, and a beating length value is then obtained. of polarization and a coupling length value and hence the value of the PMD. The main disadvantage of this method is the complexity in the treatment of pOTDR traces which requires iterative estimation of several parameters, only to obtain the PMD along the link.

Another pOTDR technique proposed in the literature is found in the article “Random-Scrambling Tunable pOTDR for Distributed Measurement of Cumulative ???? (J. Lightwave Technol., Vol 27, no. 18, pp 4164-4174, 2009). The proposed pOTDR method uses a pair of independent optical pulses, each pair consisting of two pulses with the same SOP, but with different but close wavelengths. The SOP and the central wavelength of the pair need not be correlated with the next pair. The method allows to measure the accumulated PMD in a link, however it has great complexity and high cost, since it must comprise a polarization scrambler, close wavelength lasers and acquisition and processing system. US2004 / 0084611 Al (Chen, X. et al), "Method of Evaluating PMD Fiber Using Polarization Optical Time Domain Reflectometry" describes a method based on pOTDR signal variation that allows locating high PMD portions. However, the use of pulses of constant width makes this technique limited to showing the correlation between the signal variation (VOS) and the PMD along the fiber, neither allowing to quantify the value of the fiber nor to estimate it.

OBJECTS OF THE INVENTION In view of the above, it is an object of the present invention to provide a method for: - Locating high PMD stretches in fiber optic links. - Estimate the PMD distributed in optical links. - Estimate and locate the PMD in optical links. - Provide the reading of pOTDR traces in commercial OTDR equipment; - Provide an easy-to-implement method with simplified pOTDR trace treatment to locate and estimate PMD value ranges in fiber optic links.

Brief Description of the Invention The present invention relates to a method of locating and estimating PMD in optical links based on the analysis of the degree of roughness of pOTDR trace as a result of the time-pulse width variation. The proposed method uses a simple OTDR adaptation system, which may be commercial equipment, to obtain estimated PMD value ranges and their location in fiber optic links. Such adaptation can be accomplished by passive optical components. The invention allows identifying high PMD sections and determining value ranges for said sections. The method for locating and estimating PMDs on a fiber optic link of the invention is based on light polarization phenomena, time domain optical reflectometry techniques, pOTDRs, ie polarization based OTDRs.

This method consists of launching light pulses on a fiber link under test and analyzing the roughness of the spatial distribution of optical power originating from the optical signal from Rayleigh backscattering after passing through an optical polarizing element. The location of the PMD and its magnitude are both a function of the pulse width of said OTDR and can be obtained by said roughness analysis, allowing the characterization of the analyzed fiber segment for its PMD value, by the relationship between the pulse width. and the range of possible PMD values.

Description of the Figures The invention will be better understood from the detailed description of the related figures, wherein: Figure 1 represents the behavior of the light polarization state (102) and (105) incident on two types of optical fiber. (101) and (104) with respect to the bias stroke length already known in the prior art. A high PMD fiber (101) has a stroke length (103) less than the beat length (106) of a low PMD fiber (104), i.e. the polarization state of the incident light on a fiber. High PMD repeats after a distance smaller than would be repeated for a low PMD fiber. Figure 2 is a simplified diagram of the apparatus used to validate the technique proposed in the present invention in the step of analyzing the polarization behavior of light along the analyzed fiber optic link. The apparatus consists of an OTDR (201), two optical circulating elements (203) and (204), an optical polarizing element (209) and the fiber link under test (205), already known in the prior art. The light is released from an OTDR (201) and propagates (202) and (206) through the system, undergoing Rayleigh scattering, already known in the state of the art. The backscattered light 207, 208, 210, and 211 propagates toward the launch point through the optical polarizing element 209, returning to the OTDR, and carrying information about the polarization behavior along the link. . Figure 3 shows the comparison between the resolutions obtained by launching into apparatus (201), (203), (204), (205) and (209) through an OTDR a narrow light pulse (301) and a light pulse broad (304). In the case of the narrow light pulse (301), polarization state variations (303, 305) are best acquired, since the pulse width is smaller than the beat length of the analyzed fiber (302, 306). However, for wide pulses (304), it is not possible to observe variations, since the pulse width is greater than the fiber beat length (307). Figure 4 outlines a trace of pOTDR, already described in the prior art, where one can see two stretches (401) and (403) whose PMD is larger than the central portion (402). This is possible due to the roughness present in the central stretch and absent in the other stretches, an analysis that is part of the technique object of the present invention. Figure 5 presents three types of graphs used to validate the technique object of the invention. The first (a) and the second (b) show respectively an OTDR trace (a) and a pOTDR trace (b), relating the relative amplitude of the propagating optical signal in the tested fiber link, composed of four different PMD fibers, and the distance traveled in this. The pulse width of the optical signal launched on the link is 1000ns. The third graph (c) illustrates said concept of roughness, or undulation, which will be presented below, this being the absolute value of subtracting said OTDR and pOTDR curves, the method step of the proposed invention. Also present in the graph is the roughness threshold value, which will also be conceptualized below. Figure 6 shows the same graphs as shown in Figure 5, but with pulse width of the optical signal cast on said 500ns fiber link. Figure 7 shows the same types of graphs shown in Figure 5, but with pulse width of the optical signal cast on said 100ns fiber link. Figure 8 shows the same types of graphs as in Figure 5, but with pulse width of the optical signal cast on said 50ns fiber link. Figure 9 shows the same types of graphs as in Figure 5, but with pulse width of the optical signal cast on said 20ns fiber link. Figure 10 shows the same types of graphs as in Figure 5, but with pulse width of the optical signal cast on said 10ns fiber link. Figure 11 shows the application of the windowing technique, showing one of the ways to perform it, which consists of continuously moving a window along the roughness trace and averaging the values of said window roughness. In the figure, the windowing technique is illustrated for a 20-point window, in this case, equivalent to 100m of fiber link.

Figures 1A through 11F correspond to optical signal pulses thrown at said fiber link having widths of 1000ns (Figure 11A), 500ns (Figure 11B), 100ns (Figure 11C), 50ns (Figure 11D), 20ns respectively. (Figure HE), 10ns (Figure 11F). Figure 12 presents the same application of the windowing technique illustrated in Figure 11; but with 60-point window, in this case equivalent to 300m of fiber link. Figures 12A to 12F correspond to optical signal pulses thrown at said fiber link having widths of 1000ns (Figure 12A), 500ns (Figure 12B), 100ns (Figure 12C), 50ns (Figure 12D), 20ns ( Figure 12E), 10ns (Figure 12F). Figure 13 presents the same application of the windowing technique illustrated in Figure 11, also using a 20-point window. In this case, the roughness within each window is averaged, which is offset by a value equal to its width along the pOTDR trace. Figures 13A to 13F correspond to optical signal pulses launched at said fiber link having widths of 1000ns (Figure 13A), 500ns (Figure 13B), 100ns (Figure 13C), 50ns (Figure 13D), 20ns ( Figure 13E), 10ns (Figure 13F). Figure 14 illustrates the same windowing technique applied in Figure 13, but with a 60-point window. Figures 14A through 14F correspond to optical signal pulses launched at said fiber link having widths of 1000ns (Figure 14A), 500ns (Figure 14B), 100ns (Figure 14C), 50ns (Figure 14D), 20ns ( Figure 14E), 10ns (Figure 14F).

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS The following are some preferred embodiments and uses of the PMD location and estimation method in optical links by analyzing the degree of pOTDR trace roughness as a function of pulse time width variation. released. The proposed method is based on the fact that when light propagates through a high PMD fiber segment, its polarization state (SOP) changes over the length of the fiber segment in question more rapidly. that is, at shorter distances than for a low PMD fiber stretch.

In Figure 1, a stretch 101 of high PMD fiber is illustrated; In this fiber, the propagating light has its polarization state 102 altered along the propagation distance. This change occurs because of a relationship between the PMD and a birefringence parameter, known as the bias bias length Lb as per expression (1) below: (1) where ?? is the difference between the refractive indices of the two propagation modes, one for each principal polarization state, allowed in the fiber; and ? is the wavelength of the propagating optical signal. The bias bias length Lb is the distance required for the initial polarization state of the fiber to repeat, indicated by 103 in Figure 1. The bias bias length Lb is related to the PMD by the following expression (2): ( 2) where ?? is the DGD, group differential delay in fiber, also known as first order PMD, Leo fiber length in kilometers and c is the speed of light in vacuum. Expression (2) therefore relates birefringence, expressed in picoseconds per kilometer, to the bias bias length, Lb. High PMD fibers have high birefringence and therefore Lb smaller than low PMD fibers. The variation of SOP 102 in a high PMD 101 fiber and a low PMD 104 fiber is illustrated in Figure 1. In this second case the change in bias states 105 occurs more slowly and the bias beat length Lb is larger, 106, than in the high PMD fiber. The method now proposed preferably uses the arrangement of Figure 2. A time-domain optical reflectometer-type equipment, OTDR 201, releases optical pulses of known time duration. This OTDR may be commercial equipment or specially developed for the proposed method. In particular, as will be described below, commercial equipment has time pulses of predetermined durations. Equipment specially developed for the method proposed herein may be constructed to have a greater number of temporal pulse widths, especially pulses of less than 10 ns.

According to the invention, optical pulse 202, launched by OTDR 201, arrives at the first port of a combiner 203, which is preferably an optical circulator that directs it through its second port to the first port of a second combiner 204. which is also preferably an optical circulator whose second port directs said pulse to a fiber or fiber optical link to be analyzed 205. Optical pulse 206 propagating along said link 205 is Rayleigh backscattered as known in the art. generating the counter-propagating signal 207. This counter-propagating optical pulse carries information from the entire fiber optic link 205. The counter-propagating optical pulse 207 then returns to the second port of the combiner 205 preferably an optical circulator, which this time routes it through its third. a light biasing element 209. Thus, the backscattered pulse 208, which contains polar state information The infinitesimal fiber segment from which it comes has this SOP variation information converted to optical signal amplitude variation due to its passage through the polarizing element 209.

With the proposed arrangement for the PMD location and estimation method, the backscattered Rayleigh SOP variations are converted to amplitude variations. Upon passing through said polarizing element that signal 210, it returns to the third port of the first combiner 203, preferably an optical circulator, and returns 211 to the OTDR, where it is processed, recorded and shown on screen. The arrangement described above and illustrated in Figure 2 is one embodiment of an OTDR variant known as pOTDR and the amplitude-by-distance traces are called pOTDR traces.

However, depending on the OTDR time pulse width, 201, the variation of SOP may not be detected by the receiver present in the OTDR. Figure 3 illustrates this situation: an optical pulse 301 of time width At is released on optical fiber 302. This temporal width corresponds to a pulse of length, where cjn is the speed of light on the optical fiber (n is the effective refractive index). and factor 2 comes from the fact that the optical pulse is backscattered so it travels round and round the fiber.

This length AL thus defines a spatial resolution of an OTDR. According to Figure 3, if this released pulse 301 has an AL resolution length less than the bias beating length Lb (307), the variation of SOP 303 along the fiber is detected by the OTDR receiver as an amplitude variation, roughness, or variation of the pOTDR trace from a trace obtained without the presence of polarizer 209 of Figure 2.

On the other hand, if the released pulse 304 has an AL length greater than the bias beating length Lb (307), the pulse detected in OTDR 201 has no fluctuations, as this pulse contains numerous polarization states; in other words the pulse is depolarized and the polarizer 209 of Figure 2 returns only an average of fiber 306 SOPs 305. Figure 4 illustrates a typical pOTDR trace made with the arrangement of Figure 2 with a given pulse width. In this case, smooth stretches are observed in the OTDR trace 401 and 403, and a rough stretch 402. This means that the stretches corresponding to the smooth strokes have high PMD and the rough stretch corresponds to a low PMD fiber ( where the pulse width is less than bias beating length).

Thus, using pulses of different time widths it is possible to estimate the optical link PMD and locate it in the fiber. By varying the pulse width, the transition between the roughness (roughness) and non-roughness condition allows you to demarcate an upper bound and a lower bound of a fiber stretch with a given PMD.

Specifically, the method of locating and estimating PMD in optical links comprises two stages: (i) Obtaining OTDR reference traces by either directly attaching to the OTDR 201 output the fiber link 205 of Figure 2 or by removing the polarizer 209. These traces are obtained for the different pulse widths of OTDR 201 in Figure 2 and should be as smooth as possible. Alternatively, this step can be disregarded if sufficient averages are ensured so that the pOTDR trace is not confused with noise at the end of the equipment's dynamic range and a linear adjustment is made in the rough regions to determine the roughness. (ii) Analyze the absolute value of the difference between the reference trace obtained in (i) - or by a linear fit - and the pOTDR trace. This measurement is one of the possible definitions of trace roughness or trace roughness.

If the degree of roughness or roughness in the fiber segment under analysis is greater than 0.10 dB, or other empirically determined value, then the pulse used is defined as the first pulse with PPCR roughness and the largest anterior pulse is defined as last. pulse without roughness UPSR, assuming that the traces of pOTDR are obtained from the widest to the narrowest pulse. These two PPCR and UPSR values will respectively define the upper and lower PMD limits of the stretch in question. The relationship between the OTDR pulse width 201 of Figure 2 and the birefringence value expressed in ps / km according to equation (2) is obtained by equating the OTDR optical pulse width in km, AL, and the bias stroke length, Lb through expression (3): (3) Entering expression (2) the value of Lb obtained by expression (3) gives the limit values for birefringence. In the case of a commercial OTDR, for example, the existing optical pulse values and their birefringence values in ps / km are shown in the table below: Thus, using the relationships in the table above it is possible to determine for the fiber segment. upper and lower limits of birefringence, in ps / km, related to PPCR and UPSR, respectively. By multiplying the values obtained by the length of the segment, it is possible to obtain the PMD value range, in ps, of the segment under analysis. The use of an OTDR 201, specially developed for this application, would allow the use of pulses with greater granularity temporal widths (ie, an OTDR that has different pulse widths, with different temporal lengths), which would allow to obtain ranges of values. more restrictive PMDs.

An alternative for the analysis of roughness data is the determination of the signal variation in a spatial window that runs through the analyzed trace. To do so, the width of the window is defined and a maximum value, an average value, the standard deviation, variance or some other characteristic that allows the roughness in each spatial window to be quantified are determined in that interval. Thus, a graph of the variation of the signal related to the degree of roughness as a function of distance is obtained. Other methods for roughness analysis may be adopted that are variations of this idea.

As an example of experimental validation of the proposed method, the arrangement of Figure 2 was tested on a link consisting of four fiber types with different PMD. The table below shows the characteristics of each fiber, spliced in sequence and designated as FIBER 1, FIBER 2, FIBER 3 and FIBER 4. The characteristics of each fiber were determined by a commercial OTDR and a commercial PMD meter. below: In the table above, the length and PMD values are measured with commercial equipment and the birefringence determined by the ratio between the fiber PMD and its length.

Tests using the proposed method were performed in order to validate the present invention using the following optical pulse widths: 1000ns, 500ns, 100ns, 50ns, 20ns and 10ns. The results obtained will be commented below.

The results shown in Figure 5a and Figure 5b show, for a light pulse width of 1000ns, the OTDR (Figure 5a) and pOTDR (Figure 5b) curves obtained by the object technique of the invention. Figure 5c illustrates the comparison between the analyzed fiber link roughness curve and the empirically defined threshold curve.

It is noted that only FIBRA 1 presented roughness higher than the threshold, so, according to the proposed method, the 1000ns pulse is the first roughness pulse, PPCR, information that will be used to estimate its PMD value. . The other fibers, FIBRA 2, FIBRA 3 and FIBRA 4 present roughness below the threshold value, which concludes that their PMDs have a higher value than the FIBRA 1 PMD.

The results shown in Figure 6a and Figure 6b show, for a light pulse width of 500ns, the OTDR (Figure 6a) and pOTDR (Figure 6b) curves obtained by the technique object of the invention. Figure 6c illustrates the comparison between the analyzed fiber link roughness curve and the empirically defined threshold curve. No significant changes were observed at this stage of the test, and it was found that the FBP 2, 3 and 4 PMD is in a greater range than the 500ns pulse emission.

The results shown in Figure 7a and Figure 7b show, for a light pulse width of 100ns, the OTDR (Figure 7a) and pOTDR (Figure 7b) curves obtained by the technique object of the invention and Figure 7c illustrates the comparison. between the roughness curve of the analyzed fiber link and the empirically defined threshold curve. In this phase of the test, no significant changes were observed either, indicating that the PMD of FIBERS 2, 3 and 4 is in a larger value range than the 100ns pulse emission.

The results shown in Figure 8a and Figure 8b show, for a 50ns light pulse width, the OTDR (Figure 8a) and pOTDR (Figure 8b) curves obtained by the technique object of the invention and Figure 8c illustrates the comparison. between the roughness curve of the analyzed fiber link and the empirically defined threshold curve. Unlike the results obtained for larger pulse widths, FIBRA 4 presented roughness greater than the threshold, so, according to the proposed method, the 50ns pulse is the first pulse with roughness, PPCR, and the 100ns pulse is the last non-roughening pulse, UPSR, which information will be used to estimate your PMD value. The other fibers, FIBRA 2 and FIBRA 3 still present roughness below the threshold value, which concludes that their PMDs have a higher value than the FIBRA 1 and FIBRA 4 PMDs.

The results shown in Figure 9a and Figure 9b show, for a light pulse width of 20ns, the OTDR (Figure 9a) and pOTDR (Figure 9b) curves obtained by the technique object of the invention and Figure 9c illustrates the comparison. between the roughness curve of the analyzed fiber link and the empirically defined threshold curve. No significant changes were observed at this stage of the test, indicating that the FBP 2 and 3 PMD is in a greater range than the 20ns pulse emission.

The results shown in Figure 10a and Figure 10b show, for a light pulse width of 10ns, the OTDR (Figure 10a) and pOTDR (Figure 10b) curves obtained by the technique object of the invention and Figure 10c illustrates the comparison. between the roughness curve of the analyzed fiber link and the empirically defined threshold curve. Unlike the results obtained for larger pulse widths, FIBRA 3 presented roughness greater than the threshold, so, according to the proposed method, the 10ns pulse is the first pulse with roughness, PPCR, and the 20ns pulse is the last non-roughening pulse, UPSR, which information will be used to estimate your PMD value. The FIBRA 2 roughness did not exceed the threshold value for any of the tests performed, so it can be concluded that its PMD value is higher than that of the other fibers and is in a greater range than the lowest test pulse. , with width of 10ns.

By applying the method proposed in the present invention to the data obtained and described above, it is possible to estimate the range of possible PMD values for each fiber portion of the analyzed link and to compare them with the values obtained using a commercial PMD meter. . The results are presented in the table below.

The above results confirm the validity of the method object of the present invention, validating its proposal and function of locating and estimating PMD values in a fiber link. It was even possible to locate and classify Fiber 2 as a high PMD fiber as well as to classify the analyzed link sections into PMD intervals, which substantially conform to the expected PMD values obtained by a commercial PMD meter. . In order to validate the windowing technique, already presented in the description of the figures as an alternative to the roughness data analysis, tests were performed using 20 and 60 point windows, which were used to calculate the mean with the window continuously moving. about the roughness trace. Said windowing technique can be explored and detailed mathematically by defining the concept of roughness as Ri = | Ti-Tref, i |, such that Ti is the ith point of the pOTDR trace and Tref, i is the ith point of the conventional OTDR or pOTDR reference trace by removing the polarizer. The window itself takes place using the average of several roughness measurements Ri according to the number of points of the window, where N is the said number of points.

In the case of the proposed technique, the values N = 20 points and N = 60 points were used and for each of these cases the window of N points can be slid each N points so that there is no covering between points already used in the averaging the window to move at each point in order to cover the points already used and recalculate the average of the roughness values, making the analyzed trace smoother and less noisy.

The results of said tests are shown in Figure 11 and Figure 12, respectively. This analysis was also performed by displacing the 20-point and 60-point windows non-continuously, but with equal size displacement of the window dimension whose results are shown in Figure 13 and Figure 14, respectively.

Comparing the results obtained by the said windowing technique with the results presented in the table above, it is possible to verify the precision of the windowing method with respect to the level of roughness of each fiber stretch of said link analyzed for each width of said window. pulse thrown.

While the present invention has been described in connection with certain preferred embodiments, it should be understood that it is not intended to be limited to those particular embodiments. Rather, it is intended to cover all possible alternatives, modifications and equivalents within the spirit and scope of the invention.

Claims (4)

1. METHOD OF LOCALIZATION AND ESTIMATION OF PMD IN OPTICAL LINES, characterized by understanding the analysis of the degree of roughness of pOTDR traces obtained from the emission by an OTDR (201) of a pulse sequence (301, 304 ) of various time widths, routed to a fiber optic link (205). Method according to claim 1, characterized in that it comprises the release of light pulses on a fiber link (205) under test and comparing, with a predefined threshold level, the roughness of the pOTDR trace corresponding to the backscattered signal. (207, 208, 210, 211) By Rayleigh effect on said link (205), said backscattered signal path comprising an optical polarizer (209). Method according to claim 2, characterized in that said pulses are ordered in a sequence ranging from the largest time-width pulses (304) to the smallest time-width pulses (301). Method according to Claim 2, characterized in that said roughness is quantified by comparing the pOTDR traces in the fiber portion under analysis (205) with OTDR reference traces obtained by the removal of said optical polarizer ( 209) of the path of said backscattered signal.