JP4164070B2 - Method and apparatus for measuring polarization mode dispersion distribution in longitudinal direction of optical fiber - Google Patents

Description

  The present invention relates to an optical fiber polarization mode dispersion measuring method and measurement apparatus, and more particularly to an optical fiber to be measured using a POTDR (polarization optical time domain reflectometer) trace and a model variable of the modeled optical fiber. The present invention relates to a method and an apparatus for measuring a polarization mode dispersion distribution in the longitudinal direction of an optical fiber that calculates a group delay time difference (DGD) of the optical fiber and calculates a polarization mode dispersion (PMD) based on the group delay time difference (DGD).

  As is well known, two orthogonal modes propagate in an optical fiber, but the polarization state between the two modes orthogonal due to birefringence caused by the optical fiber structure not being perfectly circular. This phenomenon is called polarization mode dispersion (PMD) of an optical fiber because a phenomenon occurs in which the group velocity (group delay time difference: DGD) between two modes orthogonal to each other is different for each polarization. .

  Since PMD based on DGD of such an optical fiber is an obstacle that causes deterioration in transmission quality for recent high-speed and long-distance optical communication, the DGD is accurately measured, and PMD based on the PMD is measured. There is a need to compensate.

  In general, the PMD coefficient is obtained by normalizing the DGD measured with an optical fiber having a certain length by the optical fiber length or the square root of the optical fiber length.

  However, conventionally, various polarization mode dispersion measuring methods for optical fibers that have been implemented to measure PMD based on DGD of an optical fiber can only measure the total PMD of the fiber link. Since each part of the link cannot be identified, there are problems related to the recent increase in speed, distance, and capacity of optical communication.

  The present invention has been made in view of the above circumstances, and uses a POTDR (polarization optical time domain reflectometer) trace and a model variable of a modeled optical fiber in the longitudinal direction of the optical fiber to be measured. By enabling the output of the group delay time difference (DGD) for calculating the distribution of wave mode dispersion (PMD), the problems related to the recent increase in speed, distance, and capacity of optical communication have been solved. An object of the present invention is to provide a method and apparatus for measuring the polarization mode dispersion distribution in the longitudinal direction of an optical fiber to be obtained.

  In the present invention, a process of taking a POTDR trace and calculating a PMD is employed.

  This process will be described in detail in the embodiment with reference to the flowchart shown in FIG. 2, and an outline thereof will be described here.

  In the POTDR trace, first, the POTDR trace is recorded while paying sufficient attention to the width and resolution of the OTDR pulse of a certain wavelength and the spectrum of the OTDR pulse light source.

  The POTDR trace is then converted to a polarization signal to remove attenuation in the optical fiber.

  The power spectrum of this polarization signal is calculated using a Fourier transform, and the measured power spectrum of the polarization signal is fitted to an optical fiber model.

  In this case, there are many possible models, which can be extracted as parameters related to the linear birefringence component and the circular birefringence component and the coupling length.

  Such a series of processes is repeated at other wavelengths, and the wavelength dependence of these parameters is obtained.

  Finally, the PMD of the optical fiber is calculated by returning the wavelength dependent parameter to the model.

Specifically, according to the present invention, in order to solve the above problem,
(1) A parallel polarization component and a vertical polarization component in backscattered light returning to one end of the optical fiber for measurement by entering an optical pulse having a predetermined wavelength into one end of the optical fiber for measurement; A first step of outputting a polarization signal related to the optical fiber to be measured including at least one of:
A second step of outputting, by simulation, a polarization signal for the modeled optical fiber based on a predetermined model variable of the modeled optical fiber;
A third step of determining a degree of fit between both the polarization signal related to the optical fiber to be measured and the polarization signal related to the modeled optical fiber;
When a desired fit between the two signals cannot be obtained under the predetermined wavelength, until a predetermined model variable of the modeled optical fiber is adjusted to optimize the fit between the two signals. A fourth step repeating the second and third steps;
When a desired degree of fit between the two signals is obtained under the predetermined wavelength, the first to fourth steps are repeated by changing the wavelength of the optical pulse incident on one end of the optical fiber to be measured, A fifth step of calculating a model variable having wavelength dependence of the modeled optical fiber;
The group delay time difference (DGD) of the optical fiber for measurement is calculated based on the model variable having the wavelength dependence, and the polarization of the optical fiber for measurement is calculated based on the group delay time difference (DGD). A sixth step of calculating a mode dispersion (PMD) value;
A method for measuring the polarization mode dispersion distribution in the longitudinal direction of the optical fiber is provided.

Further, according to the present invention, in order to solve the above problems,
(2) The predetermined model variables of the modeled optical fiber include an optical fiber coupling length Lc as an irregular component σ, a linear birefringence component δβl, and a circular birefringence component δβc. The model variables having the wavelength dependence of the optical fiber include the coupling length Lc (λ) of the optical fiber as the irregular component σ, the linear birefringence component δβl (λ), and the circular birefringence component δβc (λ). The method of measuring the polarization mode dispersion distribution in the longitudinal direction of the optical fiber according to (1) is provided.

Further, according to the present invention, in order to solve the above problems,
( 3 ) Parallel polarization component and vertical polarization component in backscattered light returning to one end of the optical fiber for measurement by entering a light pulse having a predetermined wavelength into one end of the optical fiber for measurement; A measurement unit that outputs a polarization signal related to the optical fiber to be measured including at least one of
Based on a predetermined model variable of the modeled optical fiber in advance, a calculation unit that outputs a polarization signal related to the modeled optical fiber by simulation,
A determination unit that determines a degree of fit between both the polarization signal related to the optical fiber to be measured output by the measurement unit and the polarization signal related to the modeled optical fiber output from the calculation unit; ,
When the desired fitting degree between the two signals is not obtained by the determination unit under the predetermined wavelength, a predetermined signal of the modeled optical fiber is used in order to simulate the polarization signal again in the arithmetic unit. A model variable adjustment unit for adjusting model variables;
When the desired fitting degree between the two signals is obtained by the determination unit under the predetermined wavelength, the measurement unit repeats the measurement by changing the wavelength of the light pulse incident on one end of the optical fiber to be measured. A wavelength-dependent model variable computing unit that calculates a model variable having wavelength dependence of the modeled optical fiber;
Based on the model variable having the wavelength dependence of the modeled optical fiber calculated by the wavelength dependence model variable computing unit, a group delay time difference (DGD) of the optical fiber to be measured is calculated, A polarization mode dispersion calculator for calculating a polarization mode dispersion (PMD) value of the optical fiber under measurement based on a group delay time difference (DGD);
An apparatus for measuring the polarization mode dispersion distribution in the longitudinal direction of the optical fiber is provided.

Further, according to the present invention, in order to solve the above problems,
( 4 ) The measurement unit
An optical pulse incident means for sequentially injecting an optical pulse having a predetermined wavelength among a plurality of wavelengths specified in advance to one end of the optical fiber for measurement;
When the optical pulse having the predetermined wavelength is incident by the optical pulse incident means, backscattered light returning to one end of the optical fiber for measurement is converted into at least one of a parallel polarization component and a vertical polarization component. Polarization component extracting means for extracting
Photoelectric conversion means for converting at least one of a parallel polarization component and a vertical polarization component of the backscattered light extracted by the polarization component extraction means into an electrical signal;
Based on the electrical signal converted by the photoelectric conversion means, computing means for calculating a polarization signal related to the optical fiber to be measured;
First Fourier transform means for calculating the power spectrum of the polarization signal related to the actual measurement calculated by the arithmetic means by Fourier transform;
The polarization mode dispersion distribution measuring device in the longitudinal direction of the optical fiber according to (3) is provided.

Further, according to the present invention, in order to solve the above problems,
(5) The calculation section
Model variable setting means for setting model variables including a coupling length Lc as an irregular component σ of the modeled optical fiber designated in advance, a linear birefringence component δβl, and a circular birefringence component δβc;
Simulating means for simulating the polarization signal of the modeled optical fiber based on the model variable of the modeled optical fiber preset by the model variable setting means;
Second Fourier transform means for calculating a power spectrum of the polarization signal of the modeled optical fiber simulated by the simulating means by Fourier transform;
(4), the polarization mode dispersion distribution measuring device in the longitudinal direction of the optical fiber is provided.

Further, according to the present invention, in order to solve the above problems,
( 6 ) The determination unit
The power spectrum of the polarization signal related to the optical fiber to be measured calculated by the first Fourier transform means of the measurement unit and the modeled optical fiber calculated by the second Fourier transform means of the calculation unit. A comparison means for comparing the power spectrum of the polarization signal;
Fit degree determination means for determining a fit degree between the two polarization signals based on the comparison result compared by the comparison means;
The polarization mode dispersion distribution measuring device in the longitudinal direction of the optical fiber according to (5) is provided.

Further, according to the present invention, in order to solve the above problems,
( 7 ) The model variable setting means preliminarily stores the coupling length Lc, the linear birefringence component δβl, and the circular birefringence component δβc as the irregular component σ of the modeled optical fiber stored in the memory. An apparatus for measuring a polarization mode dispersion distribution in the longitudinal direction of an optical fiber according to (5) is provided, wherein predetermined model variables are read out from the model variables including and supplied to the simulating means .

Further, according to the present invention, in order to solve the above problems,
( 8 ) Predetermined model variables of the modeled optical fiber include a coupling length Lc of the optical fiber as an irregular component σ, a linear birefringence component δβl, and a circular birefringence component δβc. model variable having a wavelength dependence of an optical fiber includes a coupling length of the optical fiber as irregular component sigma Lc (lambda), linear birefringence component δβl (λ), circular birefringence component δβc (λ) An apparatus for measuring a polarization mode dispersion distribution in the longitudinal direction of an optical fiber according to any one of ( 3) to (7) is provided.

  According to the present invention, a group delay for calculating polarization mode dispersion (PMD) of a fiber to be measured using a POTDR (polarization optical time domain reflectometer) trace and a model variable of a modeled optical fiber. A method for measuring the polarization mode dispersion distribution in the longitudinal direction of an optical fiber, which can solve the problems related to high speed, long distance, and large capacity of recent optical communication by enabling output of time difference (DGD), and An apparatus can be provided.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a block diagram for explaining a configuration of a main part of a longitudinal polarization mode dispersion distribution measuring apparatus for an optical fiber according to an embodiment of the present invention.

  That is, an optical fiber polarization mode dispersion measuring apparatus according to an embodiment of the present invention roughly includes a measuring unit 100, a calculating unit 200, a determining unit 300, and a PMD calculating unit 400.

  Here, the measurement unit 100 includes a POTDR trace unit 101, a polarization signal calculation unit 112, and a first Fourier (FFT) conversion unit 113 as a first Fourier conversion unit, which will be described later. ing.

  As a result, the measurement unit 100 enters a parallel pulse in the backscattered light returning to one end of the optical fiber 501 to be measured by making an optical pulse having a predetermined wavelength incident on one end of the optical fiber 501 to be measured. A polarization signal related to the optical fiber to be measured including at least one of a component and a vertical polarization component is calculated, and finally, the power spectrum is output from the first Fourier (FFT) converter 113. Has been made.

  The power spectrum of the polarization signal related to the optical fiber to be measured 501 output from the first Fourier transform unit 113 is supplied to one end of the comparison unit 301 of the determination unit 300.

  On the other hand, the calculation unit 200 includes a setting unit 201 and a model variable storage unit 202 as wavelength and model variable setting units as described later, a polarization signal simulation unit 203 as a simulation unit, and a second Fourier transform unit. And a second Fourier transform unit 204.

  Thereby, the calculation unit 200 calculates a polarization signal related to the modeled optical fiber based on a predetermined model variable of the modeled optical fiber in advance by simulation, and finally its power spectrum. Are output from the second Fourier transform unit 204.

  The power spectrum of the polarization signal relating to the modeled optical fiber from the second Fourier transform unit 204 is supplied to the other end of the comparison unit 301 of the determination unit 300.

  The comparison unit 301 includes, as comparison means, a power spectrum of a polarization signal related to the optical fiber to be measured, calculated by the first Fourier (FFT) conversion unit 113, and the second Fourier (FFT) conversion unit. The power spectrum of the polarization signal of the modeled optical fiber calculated by 204 is compared.

  The fitting degree determination unit 302 of the determination unit 300 serving as a determination unit includes the polarization signal related to the optical fiber to be measured output from the measurement unit 100 and the modeled output from the calculation unit 200. In order to determine the degree of fit between the two signals with the polarization signal relating to the optical fiber, the degree of fit between the power spectra of the two signals is determined based on the comparison result compared by the comparison unit 301. ing.

  And the method and apparatus for measuring the polarization mode dispersion distribution in the longitudinal direction of the optical fiber according to this embodiment is based on the predetermined model variable of the modeled optical fiber giving a desired degree of fit between the two signals. The polarization mode dispersion distribution in the longitudinal direction of the optical fiber to be measured is measured.

  Next, details of the measurement unit 100 and the calculation unit 200 will be described.

  First, the POTDR trace unit 101 of the measurement unit 100 has a predetermined wavelength set by the setting unit 201 from a light source 103 such as a DFB laser, for example, based on a pulse from the pulse generator / timing controller 102. An optical pulse is emitted.

  An optical pulse having a predetermined wavelength emitted from the light source 103 is amplified to a predetermined power via an erbium-doped optical fiber amplifier 104 as necessary, and then passed through an ASE filter 105 and a polarization controller 106. Thus, an optical pulse with a predetermined wavelength and polarization is formed.

  The optical pulse with the predetermined wavelength and polarization is then passed through the optical fiber circulator 107 and the polarization beam splitter 108, for example, for a device to be measured including a single mode optical fiber and a polarization maintaining optical fiber. The light is incident on one end of the optical fiber 501.

  The optical fiber circulator 107 may be an optical directional coupler or an optical switch.

  After the parallel polarization component and the vertical polarization component in the backscattered light returning to one end of the optical fiber for measurement 501 are separated through the optical fiber circulator 107 and the polarization beam splitter 108, respectively, For example, the first and second photoelectric converters 109 and 110 made of, for example, a 100-MHz PIN FET are converted into electrical signals.

  In this way, the parallel polarization component and the vertical polarization component of the backscattered light converted into the electrical signal by the first and second photoelectric converters 109 and 110 are, for example, displayed on the display 111 including an oscilloscope. Are recorded and displayed at the same time.

  In the present invention, such a form in which the parallel polarization component and the vertical polarization component of backscattered light are simultaneously recorded and displayed is referred to as a POTDR trace.

  Next, both the parallel polarization component and the vertical polarization component output from the display device 111 or at least one of them as described later are converted by the polarization signal calculation unit 112 into the light to be measured. It is calculated and output as a polarization signal for the fiber 501.

  Details of the calculation of the polarization signal related to the optical fiber 501 to be measured by the polarization signal calculation unit 112 will be described later.

  The polarization signal relating to the optical fiber to be measured 501 output from the polarization signal calculation unit 112 is used for calculation of its power spectrum by the first Fourier transform unit 113 as described above.

  Details of the calculation of the power spectrum of the polarization signal related to the optical fiber 501 to be measured by the first Fourier transform unit 113 will be described later.

  On the other hand, the setting unit 201 of the calculation unit 200 sets the wavelength of the light pulse emitted from the light source 103 of the POTDR trace unit 101 to the predetermined wavelength, and also sets the predetermined model variable of the pre-modeled optical fiber. Set.

  As the predetermined model variable of the pre-modeled optical fiber set by the setting unit 201, the modeled optical fiber irregularly stored in the model variable storage unit 202 such as a ROM in advance is stored. Predetermined model variables are read out from model variables including the coupling length Lc as the component σ, the linear birefringence component δβl, and the circular birefringence component δβc, and are supplied to the polarization signal simulator 203 serving as the simulating means. It is made to be done.

  The polarization signal simulating unit 203 has a coupling length as an irregular component σ of the modeled optical fiber as a predetermined model variable of the previously modeled optical fiber read from the model variable storage unit 202. Based on the model variables including Lc, linear birefringence component δβl, and circular birefringence component δβc, the polarization signal of the modeled optical fiber is calculated and supplied to the second Fourier transform unit 204. Has been made.

  The polarization signal relating to the modeled optical fiber from the polarization signal simulating unit 203 is used for the calculation of its power spectrum by the second Fourier transform unit 204 as described above.

  Details of the calculation of the power spectrum of the polarization signal related to the modeled optical fiber by the second Fourier transform unit 204 will be described later.

  As described above, the fit degree determination unit 302 of the determination unit 300 is the modeled signal output by the calculation unit 200 and the polarization signal related to the optical fiber 501 to be measured output by the measurement unit 100. The degree of fit between the power spectra of both signals with the polarization signal relating to the optical fiber is determined based on the comparison result compared by the comparison unit 301, and the degree of fit between the power spectra of the two signals is a desired fit. If not, the NG signal is supplied to the model variable adjustment unit 205.

  Based on the NG signal from the fitting degree determination unit 302, the model variable adjustment unit 205 is configured to generate a coupling length Lc as an irregular component σ of the modeled optical fiber, a linear birefringence component δβl, a circular birefringence. A model variable including the component δβc is adjusted and supplied to the polarization signal simulation unit 203.

  The polarization signal simulating unit 203 includes a coupling length Lc as an irregular component σ of the modeled optical fiber adjusted by the model variable adjusting unit 205, a linear birefringence component δβl, and a circular birefringence component δβc. The modeled optical fiber polarization signal is recalculated based on the model variable including and supplied to the second Fourier transform unit 204.

  The polarization signal relating to the modeled optical fiber calculated again by the polarization signal simulating unit 203 is subjected to the calculation of the power spectrum again by the second Fourier transform unit 204.

  As described above, the fit degree determination unit 302 of the determination unit 300 is the modeled signal output by the calculation unit 200 and the polarization signal related to the optical fiber 501 to be measured output by the measurement unit 100. The degree of fit between the power spectra of both polarization signals with respect to the polarization signal related to the optical fiber is determined again based on the comparison result re-compared by the comparison unit 301, and it should be the desired degree of fit. For example, the NG signal is supplied to the model variable adjustment unit 205 again.

  In this way, the adjustment of the model variable in the model variable adjustment unit 205 is repeated until the fit degree between the two polarization signals becomes a desired fit degree as a determination result in the fit degree determination unit 302. ing.

  Then, as a determination result in the fit degree determination unit 302, when the fit degree between the two polarization signals becomes a desired fit degree, an OK signal is supplied to the DGD calculation unit 401 of the PMD calculation unit 400. .

  The DGD calculation unit 401 is configured to irregularly adjust the modeled optical fiber as a model variable adjusted by the model variable adjustment unit 205 until the degree of fit between the polarization signals becomes a desired degree of fit. The DGD of the optical fiber 501 to be measured is calculated based on model variables including the coupling length Lc as the component σ, the linear birefringence component δβl, and the circular birefringence component δβc.

  The measured optical fiber based on model variables including a coupling length Lc as an irregular component σ of the modeled optical fiber by the DGD computing unit 401, a linear birefringence component δβl, and a circular birefringence component δβc. Details of the calculation of the DGD 501 will be described later.

  Thereby, the DGD calculating part 401 can output the information regarding the phase birefringence of the optical fiber 501 to be measured.

  However, when performing PMD calculation, the wavelength of the optical pulse emitted from the light source 103 of the POTDR trace unit 101 differs from the predetermined wavelength by the wavelength setting of the setting unit 201 of the calculation unit 200. Are emitted sequentially from the light source 103 of the POTDR trace unit 101.

  As a result, the measurement unit 100, the calculation unit 200, and the determination unit 300 perform measurement, calculation, and determination as described above on the wavelength of an optical pulse having some other wavelengths different from the predetermined wavelength. It is made to repeat correspondingly.

  As a result, the coupling length Lc as the irregular component σ of the modeled optical fiber, the linear birefringence component δβl, the circular birefringence is finally obtained by the wavelength dependent model variable calculation unit 206 of the calculation unit 200. The model variables including the component δβc are respectively the coupling length Lc (λ) as the irregular component σ (λ) of the modeled optical fiber having wavelength dependence, the linear birefringence component δβl (λ), Calculated as a circular birefringence component δβc (λ).

  A coupling length Lc (λ) as an irregular component σ (λ) of the modeled optical fiber having wavelength dependence by the wavelength dependence model variable calculation unit 206, a linear birefringence component δβl (λ), Details of the calculation of the circular birefringence component δβc (λ) will be described later.

  Then, the DGD calculation unit 401 corresponds to the wavelength of an optical pulse having some other wavelength different from the predetermined wavelength, and the fitting degree between the two polarization signals becomes a desired fitting degree. Until now, the coupling length as the irregular component σ (λ) of the modeled optical fiber adjusted by the model variable adjusting unit 205 and having the wavelength dependency calculated by the wavelength dependency model variable calculating unit 206 Based on the model variables including Lc (λ), linear birefringence component δβl (λ), and circular birefringence component δβc (λ), the DGD of the optical fiber 501 to be measured is calculated, and the PMD calculation unit 400 The PMD calculation processing unit 402 is supplied.

  A coupling length Lc (λ) as an irregular component σ (λ) of the modeled optical fiber having wavelength dependency by the DGD calculation unit 401, a linear birefringence component δβl (λ), a circular birefringence component Details of the calculation of the DGD of the optical fiber 501 to be measured based on the model variables including δβc (λ) will be described later.

  Then, the PMD arithmetic processing unit 402 has a coupling length Lc (λ as an irregular component σ (λ) of the modeled optical fiber having the wavelength dependency calculated by the wavelength dependency model variable calculating unit 206. ), Linear birefringence component δβl (λ), circular birefringence component δβc (λ), and based on the DGD of the optical fiber 501 to be measured calculated by the DGD calculation unit 401 based on the model variable. Predetermined PMD calculation processing of the measurement optical fiber 501 is performed.

  Details of the predetermined PMD calculation processing of the optical fiber for measurement 501 based on the DGD of the optical fiber for measurement 501 by the PMD arithmetic processing unit 402 will be described later.

  FIG. 2 is a flowchart for explaining the operation of the main part of the optical fiber polarization mode dispersion (PMD) measuring apparatus and method according to the embodiment of the present invention configured as described above.

  That is, when the start of the measurement operation is instructed, the POTDR trace by the POTDR trace unit 101 of the measurement unit 100 as described above is recorded and displayed on the display unit 111 in step S1.

  In this POTDR trace, as described above, a light pulse having a predetermined wavelength is incident on one end of the optical fiber 501 to be measured, and the parallel polarization of the backscattered light returning to one end of the optical fiber 501 to be measured. The wave component and the vertical polarization component are simultaneously recorded and displayed.

  In this case, the effective POTDR pulse width determined by the actual optical pulse width and the impulse response of the first and second photoelectric converters 109 and 110 as detectors are sufficient to elucidate the details of the POTDR trace. Must.

  Therefore, the spectral width of the light pulse from the POTDR light source must be narrow enough to prevent dipolarization and wide enough to avoid coherent Rayleigh scattering noise.

That is, it is necessary to satisfy the relationship represented by the following formula (1).

  Where Δτ is the total system PMD, λ is the operating wavelength of POTDR, Δλ is the spectral width from the POTDR light source (unit is wavelength), C is the speed of light, Δν is the spectral width from the POTDR light source (unit is the frequency), Δt Is the POTDR pulse width.

  3 and 4 show a part of each measurement example of the parallel POTDR trace and the vertical POTDR trace.

  Next, in step S2, based on the parallel POTDR trace and the vertical POTDR trace results recorded and displayed on the display unit 111 as described above, the polarization signal calculation unit 112 uses the polarization related to the optical fiber 501 to be measured. A wave signal is calculated.

  In this case, all DC offsets must first be removed from the two traces so that the zero signal corresponds to a zero backscattered light intensity.

  The sum of the parallel and vertical traces is equal to normal unpolarized OTDR. Thus, the two traces are summed to adjust the differential gain and delay between the two detectors to minimize noise on the summed trace.

  In the calculation of the polarization signal, the polarization signal is calculated in consideration of zero offset, differential gain, and delay.

  FIG. 5 shows an example of the calculated polarization signal. In order to perform sufficiently reliable measurement, the polarization signal is changed from +1 to −1 as shown in FIG.

  In determining the DC offset, the trace portion shown before the POTDR pulse enters the optical fiber 501 to be measured as a fiber system is used as the reference zero level.

  FIG. 6 shows an example of a POTDR trace with area marks. The average level of the two regions 1 (OP) and 2 (OO) marked in FIG. 6 is set to zero offset.

  Region 1 (OP) is a parallel POTDR trace, and region 2 (OO) is a vertical POTDR trace.

  When obtaining the differential gain and the delay, the path is long and the insertion loss of the first photoelectric converter 109 as a parallel polarization detector is also different. Therefore, the first and second photoelectric detectors as two detectors are different. It should be noted that there is a difference in signal level and position from the converters 109 and 110.

  This can be described as a differential gain ΔG and a delay ΔZ. These two parameters are obtained by modifying the parallel and vertical POTDR traces to minimize the total gain noise and delay.

The total trace S (z) is given by the following equation (2).

Where Sp (z) is the measured parallel polarization trace,
So (z) is the measured vertical polarization trace.

  The S (z) of the uniform fiber section exhibits an exponential delay, and the logarithm of S (z) matches a straight line using linear regression.

This is given by the following equation (3).

Where a is fiber attenuation,
b is a constant,
ε (z) is an error between the measured value at z and linear regression.

  The RMS noise is calculated as the RMS of ε (z) in the entire measurement unit 100.

The differential gain ΔG and delay ΔZ are adjusted so that this RMS noise is minimized.

In calculating the polarization signal, the polarization signal P (z) is calculated using the following equation (4) from the zero offset, the differential gain, and the delay obtained as described above.

  In the above, the configuration and calculation of the optical directional coupler 107 and the polarization beam splitter 108 as the polarization separation means of the measurement unit 100 and the first and second photoelectric converters 109 and 110 as the POTDR detectors are described. In order to simplify the above, only one of parallel and vertically polarized waves may be used.

  Thus, when only one POTDR can be used, the polarization signal can be calculated using a local maximum signal in consideration of attenuation in the fiber.

  This local maximum signal is proportional to the local average signal, and the proportionality constant is determined from the relationship between the polarization state of emission and reception.

  And even if only one parallel POTDR trace is available, the zero offset OP is calculated as described above.

  For this parallel POTDR trace only, the local maximum signal level is 1.5 times the local average signal level.

  In the uniform fiber section, the local average signal level is known to exhibit exponential attenuation.

Then, the RMS error ε (z) is minimized by applying the non-linear least square method as shown in the following equation (5).

  Execution of linear regression is not possible in the logarithmic domain, because Sp (z) varies greatly and a fairly weighted fit is made.

In the case of only this parallel POTDR trace, the polarization signal is calculated using the following equation (6).

  Also, even when only one vertical POTDR trace is available, the zero offset OO is calculated by the method of clause as described above.

  For this vertical POTDR trace only, the local maximum signal level is three times the local average signal level.

  In the uniform fiber section, the local average signal level exhibits exponential decay.

By applying a non-linear least square method as shown in the following equation (7), the RMS error ε (z) is minimized.

  Execution of linear regression is not possible in the logarithmic domain, because So (z) fluctuates so much that a fairly weighted fit is made.

In the case of only this vertical POTDR trace, the polarization signal is calculated using the following equation (8).

  Next, in step S4, the polarization power spectrum is calculated by fast Fourier transform (FFT) using the polarization signal calculated as described above.

  In this case, the power spectrum of the polarization signal is calculated by obtaining the square of the magnitude of the result of the fast Fourier transform of the polarization signal.

  Then, the polarization signal of the section of the trace for calculating PMD is extracted from the entire polarization signal.

For ease of computation, a fast Fourier transform algorithm can be used if 2 ^N point length sections are taken.

  FIG. 7 shows an example of the polarization power spectrum Fm (υ) calculated in this way.

  Compensation for the limited POTDR resolution can be performed at this stage.

  This is accomplished by dividing the calculated polarization power spectrum by the response power spectrum of the device.

  This is equivalent to removing the renormalization of the device response from the POTDR trace.

Such removal of renormalization may cause a large increase in noise in the high-frequency part, so care must be taken.

  Next, a series of processes for simulating the polarization signal in steps S5, S6, and S7 is performed.

  First, the outline of these processes will be described. A polarization signal is simulated by simulating birefringence in an optical fiber.

  As this simulation, for example, the simulation is performed by performing a Monte Carlo simulation.

  In this case, the fiber is divided into short birefringent elements, and the birefringence of each element is generated randomly according to the birefringence model.

  Then, the Mueller matrix of each element is calculated from the birefringence vector, and the Mueller rotation matrix of propagation from the first element to the nth element is calculated from multiplication of the Mueller rotation matrix for n elements.

  In addition, an output polarization signal for bi-directional transmission passing through n element fibers is calculated.

FIG. 8 shows the arrangement of the birefringent elements.

In FIG. 8, the birefringence vector of the i-th element is denoted by b (i), the forward Mueller rotation matrix of the i-th element is R (i), and the backward transmission Mueller rotation matrix of the i-th element is of the forward matrix. Transpose R (i) ^T

  The input polarization state is indicated by si, the output polarization state is indicated by so, and the backscattering rotation matrix is indicated by symbol M.

  Next, setting of model parameters in step S5 will be described. As the birefringence model, several birefringence models can be used.

  Each model has parameters indicating the intensity of three input parameters, a linear birefringence component δβl, a circular birefringence component δβc, and an irregular component σ.

  The polarization mode coupling parameter is related to the parameter σ. Each model has many random variables ν, which are irregular elements in the Monte Carlo simulation.

  For each birefringent element, these values are chosen randomly from a Gaussian distribution with zero mean and unit standard deviation.

  The length of each element is given by δl, which is equal to the sample interval of the POTDR trace for ease of calculation.

  Next, a weak perturbation model will be described. The birefringence of each element is considered to be constant for small perturbations.

The birefringence vector of the i-th element b (i) is expressed by the following equation (9).

Where δβl is a linear component of birefringence,
δβc is a circular component of birefringence,
σ is the strength of the perturbation,
δl is the element length,
νi is an independent irregular value selected from the Gaussian distribution with zero mean and unit standard deviation.

  Next, the rotation model will be described. The birefringence of each element is indicated by a linear birefringence component whose axis changes irregularly and a circular birefringence component whose intensity changes irregularly.

The birefringence vector of the i-th element b (i) is expressed by the following equation (10).

Where δβl is a linear birefringence component,
δβc is a circular birefringence component.

Θ (i) and φ (i) are given by the following relational expression (11).

Where σ is the strength of rotation rate,
δ1 is the length of the element,
νθ and νφ are independent irregular variables selected from the Gaussian distribution with zero mean and unit standard deviation.

  Next, describing the diffusion model, the diffusion process describes the birefringence of each element.

The birefringence vector of the i-th element b (i) is expressed by the following equation (12).

Where δβl is a linear component of birefringence,
δβc is a birefringent circular component,
δl is the element length,
α is the diffusion constant,
νi is an independent irregular value selected from the Gaussian distribution with zero mean and unit standard deviation.

  Next, the calculation of the Mueller rotation matrix will be described. The Mueller rotation matrix of each element is calculated from the birefringence vector of each element.

Using the following equations (13), (14), (15), three rotation matrices, Rx, Ry and Rz are calculated for each element.

By combining these, the Mueller rotation matrix R (i) of the i-th element is obtained by the following equation (16).

Next, calculation of the backscattering polarization state will be described. The polarization state of the backscattering signal received from the point immediately after the nth element is given by the following equation (17).

However, M is a symmetric matrix given by the following equation (18).

Next, the calculation of the polarization signal in step S6 will be described. First, the emission polarization state s _i is set as in the following equation (19).

  The polarization signal calculated as described above is simulated by the first element in the output polarization state.

This is equivalent to taking the upper left element of the combined rotation matrix, and the polarization signal P (n) by n elements is given by the following equation (20).

  FIG. 9 shows an example of a simulated polarization signal.

Next, the calculation of the power spectrum of the simulated polarization signal in step S7 will be described. The power spectrum of the simulated polarization signal is calculated as described above. As in the case of signals, the calculation is performed by a method using fast Fourier transform (FFT).

  The simulated polarization signal is then repeated multiple times to eliminate the noise in the simulated polarization power spectrum, and the power spectrum of each simulation is averaged to obtain the power spectrum of the simulated polarization signal. Fs (υ) is obtained.

  FIG. 10 shows an example of a simulated polarization power spectrum. Next, after the polarization power spectrum simulated in step S9 is compared with the measured polarization power spectrum, model parameters are adjusted and errors are minimized in step S10.

  First, a comparison between the simulated polarization power spectrum and the measured polarization power spectrum in step S9 will be described.

  That is, in step S9, the simulated polarization power spectrum is compared with the measured polarization power spectrum.

The fit by this comparison can be calculated from the sum of squared deviations between the two normalized traces for the appropriate frequency window defined by equation (21).

Where υ1 and υ2 are not affected by DC and are within a suitable frequency range below the high frequency noise,
Fm (υ) is the measured polarization power spectrum,
Fs (υ) is a simulated polarization power spectrum.

The degree of fit becomes the minimum value when the best fit is achieved.

Next, the adjustment of the model parameter and the error minimization in step S10 will be described. The model parameter is adjusted until Q becomes a minimum value.

  Although this method requires considerable work to do this quickly and reliably, one possible method is the simulated polarization power spectrum Fs for various values of δβl, δβc and σ. (Υ) is calculated to calculate Q as a function of δβl, δβc and σ.

  The best fitting parameters are obtained as δβl, δβc and σ that minimize this Q.

  Next, the measurement by another wavelength in step S11 will be described.

Steps S1 to S10 described above describe the procedure for calculating δβl, δβc and σ at a single wavelength. From this, information on phase birefringence about the optical fiber 501 to be measured is obtained.

  However, when calculating the PMD for the optical fiber 501 to be measured, since this PMD is a model parameter (with time delay and no phase delay), the simulation parameters, δβl, δβc, and σ depend on the wavelength. Sex is necessary.

  Therefore, in step S11, the above-described processing from step S1 to step S10 is repeated by performing measurement at another wavelength.

  Thus, the model parameters δβl (ω), δβc (ω) and σ (ω), that is, δβl (λ), δβc (λ), and σ (λ) depending on the optical frequency, that is, the wavelength, are calculated.

  There can be many models for these parameters that depend on the optical frequency or wavelength, but the σ associated with the coupling length of the polarization mode is considered to be largely independent of the optical frequency or wavelength.

  The model that may depend on the optical frequency, that is, wavelength, of the birefringence components δβl and δβc is a combination of phase delay and mode delay with a constant multiplier.

This is expressed as the following equation (22).

Thus, a linear birefringence component depending on the optical frequency (that is, wavelength) shown in the following formula (23) is obtained.

  A similar equation can be used for the circular birefringence component.

X _l is between 1 and 2, Xc is expected to be about 0.1.

  Next, the calculation of PMD in step S12 will be described.

First, the output polarization state depending on the optical frequency, that is, the wavelength is calculated from the fiber section of the optical fiber 501 to be measured using the parameters of the model that depends on the optical frequency, that is, the wavelength.

  According to the method as described above, but only the one-way output state is calculated as a function of the optical frequency ω.

In this way, s _o (ω) is given by the following equation (24).

Since the Mueller rotation matrix R (ω) for the fiber section has been calculated, the differential group delay is determined by the optical frequency, for example, according to known techniques well described by Non-Patent Documents 1 to 3 below. Is calculated as a function of
C. D. POOL, N.M. S. BERGANO, R.A. E. WAGNER, H.W. J. et al. SCHULTE, “Polarization Dispersion and Principal States in a 147 km Undersea Lightwave Cable” J of Light Tech. , Vol 6, No7, July 1988, pp 1185-1190. B. L. HEFFNER, "Automated Measurement of Polarization Mode Dispersion Using Jones Matrix Genetics", IEEE Photonics Tech Lett, Vol4, No9, Sep. 9106. L. E. NELSON, R.M. M.M. JOPSON, H.M. KOGELNIK, "Muller Matrix Method for Determining Polarization Mode Disposition Vectors", ECOC 1999, Nice Francece, Wednesday 29 September 1999, Sess. The PMD is then calculated as the RMS (RMS) DGD or the average value DGD depending on the definition required.

  Such simulation can be repeated several times to obtain a better estimated value of PMD.

FIG. 1 is a block diagram for explaining a configuration of a main part of a longitudinal polarization mode dispersion distribution measuring apparatus for an optical fiber according to an embodiment of the present invention. FIG. 2 is a flowchart for explaining the operation of the main part of the apparatus and method for measuring the polarization mode dispersion distribution in the longitudinal direction of the optical fiber according to the embodiment of the present invention configured as described above. FIG. 3 is a diagram showing a part of a measurement example of a parallel POTDR trace. FIG. 4 is a diagram showing a part of a measurement example of a vertical POTDR trace. FIG. 5 is a diagram illustrating an example of the calculated polarization signal. FIG. 6 is a diagram illustrating an example of a POTDR trace with an area mark. FIG. 7 is a diagram illustrating an example of the calculated polarization power spectrum Fm (υ). FIG. 8 is a diagram showing the arrangement of birefringent elements. FIG. 9 is a diagram illustrating an example of a simulated polarization signal. FIG. 10 is a diagram illustrating a simulated polarization power spectrum.

Explanation of symbols

100 ... measurement part,
200 ... arithmetic unit,
300 ... determination part,
400 ... PMD calculation unit,
401 ... DGD calculation unit,
402: PMD arithmetic processing unit,
501 ... Optical fiber for measurement,
101 ... POTDR trace section,
112 ... Polarization signal calculation unit,
113... First Fourier transform means (first Fourier transform unit),
301 ... Comparison part,
302 ... fitting degree determination unit,
201 ... setting unit,
202 ... Model variable storage unit,
203 ... Simulating means (polarization signal simulating unit)
204: Second Fourier transform means (second Fourier transform unit)
205 ... model variable adjustment unit,
102 ... Pulse generator / timing controller,
103 ... light source,
104: Erbium-doped optical fiber amplifier,
105 ... ASE filter,
106: Polarization controller,
107: optical fiber circulator,
108: Polarization beam splitter,
109 ... the first photoelectric converter,
110 ... second photoelectric converter,
111: Display.

Claims (8)

At least one of a parallel polarization component and a vertical polarization component in backscattered light returning to one end of the optical fiber for measurement by entering an optical pulse having a predetermined wavelength into one end of the optical fiber for measurement A first step of outputting a polarization signal relating to the optical fiber to be measured including:
A second step of outputting, by simulation, a polarization signal for the modeled optical fiber based on a predetermined model variable of the modeled optical fiber;
A third step of determining a degree of fit between both the polarization signal related to the optical fiber to be measured and the polarization signal related to the modeled optical fiber;
When a desired fit between the two signals cannot be obtained under the predetermined wavelength, until a predetermined model variable of the modeled optical fiber is adjusted to optimize the fit between the two signals. A fourth step repeating the second and third steps;
When a desired degree of fit between the two signals is obtained under the predetermined wavelength, the first to fourth steps are repeated by changing the wavelength of the optical pulse incident on one end of the optical fiber to be measured, A fifth step of calculating a model variable having wavelength dependence of the modeled optical fiber;
The group delay time difference (DGD) of the optical fiber for measurement is calculated based on the model variable having the wavelength dependence, and the polarization of the optical fiber for measurement is calculated based on the group delay time difference (DGD). A sixth step of calculating a mode dispersion (PMD) value;
A method for measuring the polarization mode dispersion distribution in the longitudinal direction of an optical fiber.   The predetermined model variables of the modeled optical fiber include a coupling length Lc of the optical fiber as an irregular component σ, a linear birefringence component δβl, and a circular birefringence component δβc, and the modeled optical fiber The model variables having the wavelength dependence include an optical fiber coupling length Lc (λ) as an irregular component σ, a linear birefringence component δβl (λ), and a circular birefringence component δβc (λ). The method of measuring the polarization mode dispersion distribution in the longitudinal direction of the optical fiber according to claim 1. At least one of a parallel polarization component and a vertical polarization component in backscattered light returning to one end of the optical fiber for measurement by entering an optical pulse having a predetermined wavelength into one end of the optical fiber for measurement A measurement unit that outputs a polarization signal related to the optical fiber to be measured, including:
Based on a predetermined model variable of the modeled optical fiber in advance, a calculation unit that outputs a polarization signal related to the modeled optical fiber by simulation,
A determination unit that determines a degree of fit between both the polarization signal related to the optical fiber to be measured output by the measurement unit and the polarization signal related to the modeled optical fiber output from the calculation unit; ,
When the desired fitting degree between the two signals is not obtained by the determination unit under the predetermined wavelength, a predetermined signal of the modeled optical fiber is used in order to simulate the polarization signal again in the arithmetic unit. A model variable adjustment unit for adjusting model variables;
When the desired fitting degree between the two signals is obtained by the determination unit under the predetermined wavelength, the measurement unit repeats the measurement by changing the wavelength of the light pulse incident on one end of the optical fiber to be measured. A wavelength-dependent model variable computing unit that calculates a model variable having wavelength dependence of the modeled optical fiber;
Based on the model variable having the wavelength dependence of the modeled optical fiber calculated by the wavelength dependence model variable computing unit, a group delay time difference (DGD) of the optical fiber to be measured is calculated, A polarization mode dispersion calculator for calculating a polarization mode dispersion (PMD) value of the optical fiber under measurement based on a group delay time difference (DGD);
An apparatus for measuring the polarization mode dispersion distribution in the longitudinal direction of an optical fiber. The measuring unit is
An optical pulse incident means for sequentially injecting an optical pulse having a predetermined wavelength among a plurality of wavelengths specified in advance to one end of the optical fiber for measurement;
When the optical pulse having the predetermined wavelength is incident by the optical pulse incident means, backscattered light returning to one end of the optical fiber for measurement is converted into at least one of a parallel polarization component and a vertical polarization component. Polarization component extracting means for extracting
Photoelectric conversion means for converting at least one of a parallel polarization component and a vertical polarization component of the backscattered light extracted by the polarization component extraction means into an electrical signal;
Based on the electrical signal converted by the photoelectric conversion means, computing means for calculating a polarization signal related to the optical fiber to be measured;
First Fourier transform means for calculating the power spectrum of the polarization signal related to the actual measurement calculated by the arithmetic means by Fourier transform;
The polarization mode dispersion distribution measuring apparatus in the longitudinal direction of the optical fiber according to claim 3, wherein The computing unit is
Model variable setting means for setting model variables including a coupling length Lc as an irregular component σ of the modeled optical fiber designated in advance, a linear birefringence component δβl, and a circular birefringence component δβc;
Simulating means for simulating the polarization signal of the modeled optical fiber based on the model variable of the modeled optical fiber preset by the model variable setting means;
Second Fourier transform means for calculating a power spectrum of the polarization signal of the modeled optical fiber simulated by the simulating means by Fourier transform;
The polarization mode dispersion distribution measuring device in the longitudinal direction of the optical fiber according to claim 4. The determination unit
The power spectrum of the polarization signal related to the optical fiber to be measured calculated by the first Fourier transform means of the measurement unit and the modeled optical fiber calculated by the second Fourier transform means of the calculation unit. A comparison means for comparing the power spectrum of the polarization signal;
Fit degree determination means for determining a fit degree between the two polarization signals based on the comparison result compared by the comparison means;
The apparatus for measuring a polarization mode dispersion distribution in the longitudinal direction of an optical fiber according to claim 5. The model variable setting means includes a model variable including a coupling length Lc as an irregular component σ of the modeled optical fiber stored in a memory in advance, a linear birefringence component δβl, and a circular birefringence component δβc. 6. The apparatus for measuring a polarization mode dispersion distribution in the longitudinal direction of an optical fiber according to claim 5, wherein a predetermined model variable is read out from the data and supplied to the simulating means. The predetermined model variables of the modeled optical fiber include a coupling length Lc of the optical fiber as an irregular component σ, a linear birefringence component δβl, and a circular birefringence component δβc, and the modeled optical fiber The model variables having the wavelength dependence include an optical fiber coupling length Lc (λ) as an irregular component σ, a linear birefringence component δβl (λ), and a circular birefringence component δβc (λ). The polarization mode dispersion distribution measuring apparatus in the longitudinal direction of the optical fiber according to any one of claims 3 to 7.

Images