A method and system for testing wavelength characteristics of isolation of an optical isolator

Abstract

The application provides a kind of optical isolator isolation wavelength characteristic test method and system, including the light path comprising tunable light source, polarization state generator and polarization state analyzer is built;When there is no device to be measured, the system calibration Jones matrix of each wavelength is measured and singular value decomposition is carried out;The forward and reverse of the isolator to be measured is accessed to the light path, and the forward and reverse hybrid Jones matrix is obtained;According to the maximum singular value of calibration matrix, the regularization threshold value is set, the calibration matrix pseudo-inverse is constructed, and the forward and reverse Jones matrix of the device is solved;The matrix element is split into mode and phase angle, a complex model is used, which includes a non-negative amplitude mode function containing a Fabry-Perot interference period term and a Cauchy dispersion slowly varying phase angle term, and a corresponding weight is fitted to obtain a continuous function of the matrix element changing with wavelength, and the isolation wavelength characteristic curve in the preset wavelength range is calculated accordingly.

Description

Technical Field

[0001] This application belongs to the field of optical isolators, and in particular relates to a method and system for testing the wavelength characteristics of the isolation of optical isolators. Background Technology

[0002] Optical isolators are core non-reciprocal devices in optical communication, fiber optic sensing, and laser measurement systems. They are used to enable forward transmission of optical signals and attenuate or block reverse echoes, preventing reflected light from causing frequency instability or damage to the light source. Isolation is a key indicator of its reverse blocking capability. With the development of optical wavelength division multiplexing and broadband optical transmission technologies, the wavelength range covered by optical signals is constantly expanding. The performance variations of optical isolators at different wavelengths directly affect system stability and transmission quality. Therefore, obtaining the characteristic curve of isolation as a function of wavelength has become an important requirement in the field of passive optical device testing.

[0003] Current methods for testing the polarization and transmission characteristics of optical isolators typically use the Jones matrix to describe the polarization response of the system and the device. To eliminate inherent system errors such as polarization-dependent loss, dispersion, and residual birefringence in the test optical path, the system calibration Jones matrix is ​​usually measured first, followed by the mixed Jones matrix after the device under test (DUT) is placed in the test. The Jones matrix of the DUT is then extracted by inverting the calibration matrix. However, in actual testing, background noise, polarization state fluctuations, and instrument measurement limitations exist. Conventional matrix inversion is prone to ill-conditioned calculations and amplifies measurement errors. Especially in reverse transmission tests, the transmitted light intensity is extremely weak, and high-frequency noise can easily drown out the useful signal. Furthermore, if a unified reference plane is not established at both ends of the DUT between the input polarization modulation optical path and the output analysis optical path, directly multiplying the overall unloaded matrix by the mixed matrix makes it difficult to accurately separate the device's own response. In addition, existing methods are mostly based on discrete wavelength point calculations, which are susceptible to single-point burst noise and cannot fully utilize the internal interference and dispersion characteristics of the device to achieve noise smoothing and trend correction. Therefore, how to suppress noise amplification in the inversion of the system's Jones matrix and obtain a continuous isolation wavelength characteristic curve by combining the broadband optical response law is an urgent problem to be solved in this field. Summary of the Invention

[0004] According to one aspect of the present invention, a method for testing the wavelength characteristics of the isolation of an optical isolator is proposed to address the problem that existing methods are easily affected by burst noise at a single wavelength point, which disrupts the continuity of the inherent interference and dispersion laws within the optical device and makes it impossible to perform noise-resistant smoothing and trend correction on the measurement data. The method includes the following steps: A test system optical path is constructed, including a tunable light source, a polarization state generator, and a polarization state analyzer. The test light output by the tunable light source within a preset wavelength range is obtained. When the optical isolator under test is not set in the test system optical path, the system calibration Jones matrix of each wavelength point is measured based on the input and output reference plane of the optical isolator under test, and singular value decomposition is performed to obtain the singular values ​​of each wavelength point. The optical isolator under test is placed into the test system optical path along the forward and reverse transmission directions, respectively, and the forward mixed Jones matrix and the reverse mixed Jones matrix of each wavelength point are measured. For each wavelength point, a regularization threshold is set based on the maximum singular value of the system calibration Jones matrix at the wavelength point. Singular values ​​smaller than the threshold are discarded to construct a pseudo-inverse matrix of the calibration Jones matrix. The pseudo-inverse matrix is ​​multiplied by the corresponding hybrid Jones matrix to calculate the forward Jones matrix and the reverse Jones matrix of the optical isolator under test at the wavelength point. The matrix elements of the forward and reverse Jones matrices are used as complex-valued functions of wavelength, and their magnitudes and phase angles are separated. A complex model function is constructed using a non-negative amplitude mode function containing a Fabry-Perot interference periodic term and a Cauchy dispersion slowly varying phase angle term. The function is fitted with a weight that decreases as the relative energy of discarded singular values ​​increases, thus obtaining a continuous function expression of the matrix elements as a function of wavelength. Based on the continuous function expression, the isolation wavelength characteristic curve of the optical isolator within a preset wavelength range is calculated.

[0005] Optionally, acquiring the test light output by the tunable light source within a preset wavelength range includes: Set the starting and ending test wavelengths of the tunable light source as the lower and upper limits of measurement, respectively. Set the step interval and tuning scan speed for wavelength scanning; Wavelength scanning control commands are sent to the tunable light source via the communication interface; The tunable light source is controlled to be tuned point by point between the lower and upper measurement limits according to the stated step interval; A continuous monochromatic test light is output at each wavelength dwell point and injected into the optical path of the test system.

[0006] Optionally, when no optical isolator under test is set in the optical path of the test system, the method of measuring the system calibration Jones matrix at each wavelength point based on the input-output reference plane of the optical isolator under test, and performing singular value decomposition to obtain the singular values ​​at each wavelength point, includes: Use an unloaded fiber optic patch cord to connect the polarization state generator and the polarization state analyzer to form a blank test optical path without the device under test; Through input-side pre-compensation, the input polarization state is equivalently defined on the input reference plane of the optical isolator under test; At each test wavelength point, the polarization state generator is controlled to output horizontally linearly polarized light, vertically linearly polarized light, 45° linearly polarized light, and right-hand circularly polarized light in sequence. The output Stokes vector data corresponding to each input polarization state is acquired using a polarization state analyzer. Based on the output Stokes vector data, the system calibration Jones matrix for the corresponding wavelength point is obtained using the Jones matrix derivation algorithm. The system calibration Jones matrix is ​​subjected to singular value decomposition, and the resulting singular values ​​are extracted as singular value parameters for the wavelength points.

[0007] Optionally, setting a regularization threshold for each wavelength point based on the maximum singular value of the system calibration Jones matrix at that wavelength point includes: Iterate through all non-zero singular values ​​obtained from the singular value decomposition of the system calibration Jones matrix at each wavelength point, extract the singular value with the largest value as the maximum singular value index, multiply the maximum singular value index by a preset scaling percentage coefficient, and use the resulting product value as the regularization threshold for lower limit interception.

[0008] Optionally, the step of using the matrix elements of the forward and reverse Jones matrices as complex-valued functions of the wavelength and separating the magnitude and phase angle, and constructing a complex model function using a non-negative amplitude magnitude function containing Fabry-Perot interferometry periodic terms and a Cauchy dispersion slowly varying phase angle term, includes: Based on the Cauchy dispersion relation and the relationship between the propagation optical path and wavelength, an expression for the slowly varying Cauchy dispersion term is constructed. Based on the periodic spectral ripple caused by multiple reflection interference, an expression for the periodic ripple term of the Fabry-Perot interference is constructed. The reference amplitude term, the slowly varying amplitude envelope term, and the Fabry-Perot interference periodic ripple term are combined to generate a non-negative amplitude mode expression. Using the Cauchy dispersion slow variable expression as the phase angle, construct a complex exponential term; Multiplying the nonnegative amplitude modulus expression by the complex exponential term yields the corresponding complex model function.

[0009] Optionally, the step of fitting the matrix elements using a decay factor that decreases as the relative energy of discarded singular values ​​increases, as weights, to obtain a continuous function expression of the matrix elements as a function of wavelength includes: Obtain the matrix element modulus and phase angle observation values ​​of the forward and reverse Jones matrices at each test wavelength point; The data fitting weights corresponding to each test wavelength point are determined based on the attenuation factor. Based on the data fitting weights, a weighted residual objective function is constructed between the predicted values ​​of the complex model function and the observed values. The model parameters of the complex model function are obtained by minimizing the weighted residual objective function. Substituting the model parameters into the complex model function generates a continuous function expression of the matrix elements as a function of wavelength.

[0010] Optionally, the step of calculating the isolation wavelength characteristic curve of the optical isolator within a preset wavelength range based on the continuous function expression includes: Substituting the target wavelength into the continuous function expressions of the forward and reverse Jones matrices, we obtain the forward and reverse Jones matrices for the corresponding wavelength point. Under unpolarized light conditions, the forward and reverse light power transmittances are calculated based on half the trace of the product of the Jones matrix and its conjugate transpose. The ratio of the forward light power transmittance to the reverse light power transmittance is used as the transmittance. The transmittance is calculated using a logarithmic transformation formula to obtain the isolation value at the corresponding wavelength point; Based on each wavelength point and its corresponding isolation value, an isolation wavelength characteristic curve of the optical isolator within a preset wavelength range is generated.

[0011] According to another aspect of the present invention, a wavelength characteristic testing system for the isolation of an optical isolator is provided, comprising the following modules: The measurement module is used to construct the optical path of the test system, which includes a tunable light source, a polarization state generator, and a polarization state analyzer. It acquires the test light output by the tunable light source within a preset wavelength range. When no optical isolator under test is set in the optical path of the test system, it measures the system calibration Jones matrix at each wavelength point based on the input and output reference plane of the optical isolator under test, and performs singular value decomposition to obtain the singular values ​​at each wavelength point. The optical isolator under test is placed into the optical path of the test system along the forward and reverse transmission directions, respectively, and the forward mixed Jones matrix and the reverse mixed Jones matrix at each wavelength point are measured and obtained. The construction module is used to set a regularization threshold based on the maximum singular value of the system calibration Jones matrix at each wavelength point, discard singular values ​​smaller than the threshold to construct a pseudo-inverse matrix of the calibration Jones matrix, and multiply the pseudo-inverse matrix with the corresponding hybrid Jones matrix to calculate the forward Jones matrix and the reverse Jones matrix of the optical isolator under test at the wavelength point. The calculation module is used to treat the matrix elements of the forward and reverse Jones matrices as complex-valued functions of wavelength and to separate the magnitude and phase angle. It constructs a complex model function using a non-negative amplitude magnitude function containing a Fabry-Perot interference periodic term and a Cauchy dispersion slowly varying phase angle term, and fits the function with a decay factor that decreases as the relative energy of discarded singular values ​​increases, to obtain a continuous function expression of the matrix elements changing with wavelength. Based on the continuous function expression, the isolation wavelength characteristic curve of the optical isolator within a preset wavelength range is calculated.

[0012] Preferably, acquiring the test light output by the tunable light source within a preset wavelength range includes: Set the starting and ending test wavelengths of the tunable light source as the lower and upper limits of measurement, respectively. Set the step interval and tuning scan speed for wavelength scanning; Wavelength scanning control commands are sent to the tunable light source via the communication interface; The tunable light source is controlled to be tuned point by point between the lower and upper measurement limits according to the stated step interval; A continuous monochromatic test light is output at each wavelength dwell point and injected into the optical path of the test system.

[0013] Preferably, when no optical isolator under test is set in the optical path of the test system, the method of measuring the system calibration Jones matrix at each wavelength point based on the input-output reference plane of the optical isolator under test, and performing singular value decomposition to obtain the singular values ​​at each wavelength point, includes: Use an unloaded fiber optic patch cord to connect the polarization state generator and the polarization state analyzer to form a blank test optical path without the device under test; Through input-side pre-compensation, the input polarization state is equivalently defined on the input reference plane of the optical isolator under test; At each test wavelength point, the polarization state generator is controlled to output horizontally linearly polarized light, vertically linearly polarized light, 45° linearly polarized light, and right-hand circularly polarized light in sequence. The output Stokes vector data corresponding to each input polarization state is acquired using a polarization state analyzer. Based on the output Stokes vector data, the system calibration Jones matrix for the corresponding wavelength point is obtained using the Jones matrix derivation algorithm. The system calibration Jones matrix is ​​subjected to singular value decomposition, and the resulting singular values ​​are extracted as singular value parameters for the wavelength points.

[0014] Preferably, setting a regularization threshold for each wavelength point based on the maximum singular value of the system calibration Jones matrix at that wavelength point includes: Iterate through all non-zero singular values ​​obtained from the singular value decomposition of the system calibration Jones matrix at each wavelength point, extract the singular value with the largest value as the maximum singular value index, multiply the maximum singular value index by a preset scaling percentage coefficient, and use the resulting product value as the regularization threshold for lower limit interception.

[0015] Preferably, the step of using the matrix elements of the positive and negative Jones matrices as complex-valued functions of the wavelength and separating the magnitude and phase angle, and constructing a complex model function using a non-negative amplitude magnitude function containing Fabry-Perot interferometry periodic terms and a Cauchy dispersion slowly varying phase angle term, includes: Based on the Cauchy dispersion relation and the relationship between the propagation optical path and wavelength, an expression for the slowly varying Cauchy dispersion term is constructed. Based on the periodic spectral ripple caused by multiple reflection interference, an expression for the periodic ripple term of the Fabry-Perot interference is constructed. The reference amplitude term, the slowly varying amplitude envelope term, and the Fabry-Perot interference periodic ripple term are combined to generate a non-negative amplitude mode expression. Using the Cauchy dispersion slow variable expression as the phase angle, construct a complex exponential term; Multiplying the nonnegative amplitude modulus expression by the complex exponential term yields the corresponding complex model function.

[0016] Preferably, the step of fitting the matrix elements using a decay factor that decreases as the relative energy of discarded singular values ​​increases, as weights, to obtain a continuous function expression of the matrix elements as a function of wavelength includes: Obtain the matrix element modulus and phase angle observation values ​​of the forward and reverse Jones matrices at each test wavelength point; The data fitting weights corresponding to each test wavelength point are determined based on the attenuation factor. Based on the data fitting weights, a weighted residual objective function is constructed between the predicted values ​​of the complex model function and the observed values. The model parameters of the complex model function are obtained by minimizing the weighted residual objective function. Substituting the model parameters into the complex model function generates a continuous function expression of the matrix elements as a function of wavelength.

[0017] Preferably, the step of calculating the isolation wavelength characteristic curve of the optical isolator within a preset wavelength range based on the continuous function expression includes: Substituting the target wavelength into the continuous function expressions of the forward and reverse Jones matrices, we obtain the forward and reverse Jones matrices for the corresponding wavelength point. Under unpolarized light conditions, the forward and reverse light power transmittances are calculated based on half the trace of the product of the Jones matrix and its conjugate transpose. The ratio of the forward light power transmittance to the reverse light power transmittance is used as the transmittance. The transmittance is calculated using a logarithmic transformation formula to obtain the isolation value at the corresponding wavelength point; Based on each wavelength point and its corresponding isolation value, an isolation wavelength characteristic curve of the optical isolator within a preset wavelength range is generated.

[0018] This invention constructs a pseudo-inverse matrix of the calibration Jones matrix through singular value decomposition and a regularization threshold set based on the maximum singular value. This reduces the error amplification caused by ill-conditioned matrix inversion and improves the calculation accuracy of the forward and reverse Jones matrices. Simultaneously, through input-side reference plane pre-compensation and output-side equivalent calibration matrix pseudo-inverse processing, the system response in the mixed Jones matrix is ​​removed under the same reference plane, avoiding the problem of difficulty in separating device responses due to direct left multiplication of the overall unloaded system matrix of the preceding and following stages by the mixed matrix. Furthermore, the matrix elements are treated as complex-valued functions of wavelength for magnitude and phase separation. A complex model is constructed using a non-negative amplitude magnitude function containing Fabry-Perot interferometry periodic terms and a Cauchy dispersion slowly varying phase angle term. An attenuation factor that decreases with increasing relative energy of discarded singular values ​​is used as the fitting weight, where the phase angle is a relative phase angle, eliminating the need to recover the global absolute optical phase. This reduces measurement noise interference, obtains a continuous functional expression of the matrix elements as a function of wavelength, and achieves high-precision measurement of the isolation characteristic curve over a wide wavelength range. Attached Figure Description

[0019] Figure 1 A flowchart of the first embodiment; Figure 2 This is a schematic diagram of the step scan wavelength mode curve; Figure 3 This is a schematic diagram of the Fabry-Perot periodic ripple component curve. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] In the first embodiment, the present invention proposes a method for testing the wavelength characteristics of the isolation of an optical isolator, such as... Figure 1 It includes the following steps: S1. Construct a test system optical path including a tunable light source, a polarization state generator, and a polarization state analyzer. Obtain the test light output by the tunable light source within a preset wavelength range. When no optical isolator under test is set in the test system optical path, measure the system calibration Jones matrix at each wavelength point based on the input and output reference plane of the optical isolator under test, and perform singular value decomposition to obtain the singular values ​​at each wavelength point. Place the optical isolator under test into the test system optical path along the forward and reverse transmission directions respectively, and measure the forward mixed Jones matrix and the reverse mixed Jones matrix at each wavelength point.

[0022] A basic optical path is constructed by sequentially connecting a tunable light source, a polarization state generator, and a polarization state analyzer using single-mode fiber. The tunable light source is controlled by a computer via a universal interface bus, performing wavelength scanning in 0.1 nm steps within a pre-set wavelength range of 1500 nm to 1600 nm. Without an optical isolator, the polarization state generator sequentially generates four independent input polarization states: horizontally linearly polarized light, vertically linearly polarized light, 45° linearly polarized light, and right-hand circularly polarized light. Input-side pre-compensation ensures that these four input polarization states are equivalently defined on the input reference plane of the device under test. The polarization state analyzer synchronously acquires the output optical Stokes parameters. Under conditions of non-depolarization or weak depolarization of the test optical path, a 2×2 complex system calibration Jones matrix for the basic optical path at each wavelength point is calculated using a Mueller matrix to Jones matrix conversion algorithm. The optical Stokes parameters are a set of intensity parameters describing the polarization state of light, typically including total light intensity, horizontal / vertical polarization difference, 45° / 135° polarization difference, and right-hand / left-hand circular polarization difference. The system calibration Jones matrix is ​​used to characterize the equivalent system response on the output side between the output reference plane of the device under test and the polarization state analyzer. When calibration is performed using an unloaded fiber optic jumper, the driving quantity of the polarization state generator is corrected by feedback of the known input polarization state, so that the influence of the input optical path on the input reference plane of the device under test is compensated to the known input state.

[0023] The calibration Jones matrix is ​​decomposed into a left singular vector matrix, a singular value diagonal matrix, and a right singular vector matrix. The singular values ​​are obtained by extracting the main diagonal elements of the singular value diagonal matrix. The optical isolator under test (OUT) is connected between the polarization state generator and the polarization state analyzer via a fiber optic connector in the forward transmission direction. The polarization state generation and measurement process is repeated to calculate the forward hybrid Jones matrix. The OUT is then reversed and connected to the optical path in the reverse transmission direction, and the same polarization state generation and measurement process is performed again to calculate the reverse hybrid Jones matrix. During both forward and reverse access measurements, the output-side connection structure, reference plane, and fiber optic connection status between the OUT and the polarization state analyzer are kept consistent. If the reverse access causes a change in the output-side connection structure, the equivalent system calibration Jones matrix of the output side in the corresponding direction is measured.

[0024] In an optional embodiment, acquiring the test light output by the tunable light source within a preset wavelength range includes: Set the starting and ending test wavelengths of the tunable light source as the lower and upper limits of measurement, respectively. Set the step interval and tuning scan speed for wavelength scanning; Wavelength scanning control commands are sent to the tunable light source via the communication interface; The tunable light source is controlled to be tuned point by point between the lower and upper measurement limits according to the stated step interval; A continuous monochromatic test light is output at each wavelength dwell point and injected into the optical path of the test system.

[0025] In the technical implementation process, the communication parameters of the tunable semiconductor laser are initialized through host computer software. The baud rate can be set to 115200 or a millisecond-level response can be achieved using a GPIB interface. The lower limit of wavelength measurement is set to 1520nm, and the upper limit is set to 1570nm. The wavelength step interval is set to 0.1nm, and the tuning scan speed is configured to 10nm / s. The parameters are sent to the tunable light source controller via the SCPI standard instruction set. After the light source starts scanning, the control program adopts a stepped step mode. That is, when the light source reaches each discrete dwell point, the wavelength tuning action is stopped and a locking mechanism is triggered. The waiting time parameter can be set to 10ms to 50ms to remove phase noise caused by transient wavelength drift. The tunable light source performs wavelength scanning in a stepped step mode, and the wavelength change curve over time is shown below. Figure 2 As shown. During this dwell period, the light source outputs a continuous monochromatic test light with a linewidth of less than 100kHz and a power stability better than 0.01dB. The output power can be set to 1mW, i.e., 0dBm, so that the light intensity injected into the optical path of the test system can meet the minimum sensitivity requirements of the polarization state analyzer and will not cause nonlinear effects.

[0026] In an optional embodiment, when no optical isolator under test is provided in the optical path of the test system, the process of measuring the system calibration Jones matrix at each wavelength point based on the input-output reference plane of the optical isolator under test, and performing singular value decomposition to obtain the singular values ​​at each wavelength point, includes: Use an unloaded fiber optic patch cord to connect the polarization state generator and the polarization state analyzer to form a blank test optical path without the device under test; Through input-side pre-compensation, the input polarization state is equivalently defined on the input reference plane of the optical isolator under test; At each test wavelength point, the polarization state generator is controlled to output horizontally linearly polarized light, vertically linearly polarized light, 45° linearly polarized light, and right-hand circularly polarized light in sequence. The output Stokes vector data corresponding to each input polarization state is acquired using a polarization state analyzer. Based on the output Stokes vector data, the system calibration Jones matrix for the corresponding wavelength point is obtained using the Jones matrix derivation algorithm. The system calibration Jones matrix is ​​subjected to singular value decomposition, and the resulting singular values ​​are extracted as singular value parameters for the wavelength points.

[0027] During system calibration, a 1m to 2m long polarization-maintaining or single-mode unloaded fiber optic patch cord can be used, with APC connectors at both ends to control the end-face return loss below -60dB. This connects the polarization state generator and the polarization state analyzer, constructing a reference test optical path without isolators. The host computer sends four sets of defined voltage drive signals sequentially to the polarization state generator via the control bus. These signals are then sent to the electro-optic phase delay unit inside the polarization state generator, generating Stokes parameter sequences of [1,1,0,0] for horizontal linear polarization, [1,-1,0,0] for vertical linear polarization, [1,0,1,0] for 45° linear polarization, and [1,0,0,1] for right-hand circular polarization at each set wavelength dwell point. These four input polarization states are consistent with the four input polarization states used previously, forming the independent polarization input conditions required to solve the Jones matrix. When the optical signals of these four known polarization states sequentially enter the polarization state analyzer, the detector array inside the analyzer acquires the output Stokes vector matrix at a sampling rate of 1MHz for the corresponding state. Using the standard Jones matrix derivation algorithm, a 2×2 complex matrix representing the system's background transmission characteristics at a given wavelength is constructed through the mapping relationship from the Mueller matrix to the Jones matrix. This mapping relationship is performed under the condition that the test optical path satisfies either non-depolarization or weak depolarization. When the measured Mueller matrix contains a weak depolarization component, its closest non-depolarization matrix component is extracted first, and then the corresponding equivalent Jones matrix is ​​calculated. In this embodiment, the phase of the Jones matrix derived from the Stokes vector is represented by the relative phase at the reference matrix element or reference wavelength, without prioritizing the recovery of the global absolute optical phase. To evaluate the condition number and noise immunity of the calibration matrix, a mathematical computation library is used to perform singular value decomposition on the matrix, extracting the main diagonal elements of the middle diagonal matrix. For example, the values ​​calculated at 1550nm are 0.98 and 0.96. These two dimensionless, non-negative real values ​​are temporarily stored in the memory database as characteristic parameters for calculating the system's transmission loss and polarization-dependent loss.

[0028] S2, for each wavelength point, a regularization threshold is set based on the maximum singular value of the system calibration Jones matrix at the wavelength point, and singular values ​​smaller than the threshold are discarded to construct a pseudo-inverse matrix of the calibration Jones matrix. The pseudo-inverse matrix is ​​multiplied by the corresponding hybrid Jones matrix to calculate the forward Jones matrix and the reverse Jones matrix of the optical isolator under test at the wavelength point.

[0029] Extract the two singular values ​​from the system calibration Jones matrix obtained in the preceding steps, and select the larger one as the largest singular value. Multiply this largest singular value by a preset scaling percentage factor as the regularization threshold. The scaling percentage factor can be set between 0.5% and 5% based on the signal-to-noise ratio of the test system, and can be 1.5% for example. Compare the two singular values ​​with the regularization threshold. If the smaller singular value is less than the threshold, set the diagonal element corresponding to the smaller singular value in the singular value diagonal matrix to 0, constructing a truncated singular value diagonal matrix. Take the reciprocal of the non-zero diagonal elements in the truncated singular value diagonal matrix to obtain the inverse singular value diagonal matrix. If the system calibration Jones matrix is ​​represented as... Then, take the reciprocal of the retained singular values ​​and construct... Set the corresponding position of the truncated singular value to zero, according to The pseudo-inverse of the calibration Jones matrix at that wavelength is calculated. For each wavelength, the pseudo-inverse is multiplied by the forward and reverse mixed Jones matrices at that wavelength to obtain the true forward and reverse Jones matrices of the optical isolator after removing the polarization effect of the test system. Since the input polarization state has been equivalently defined on the input reference plane of the device under test through input-side pre-compensation in S1, both the forward and reverse mixed Jones matrices can be expressed as the cascade result of the output-side system calibration Jones matrix and the Jones matrix of the optical isolator under test. Therefore, multiplying the pseudo-inverse of the output-side system calibration Jones matrix by the mixed Jones matrix on the left can achieve device response separation under the same reference plane.

[0030] In an optional embodiment, setting a regularization threshold for each wavelength point based on the maximum singular value of the system-calibrated Jones matrix at that wavelength point includes: Iterate through all non-zero singular values ​​obtained from the singular value decomposition of the system calibration Jones matrix at each wavelength point, extract the singular value with the largest value as the maximum singular value index, multiply the maximum singular value index by a preset scaling percentage coefficient, and use the resulting product value as the regularization threshold for lower limit interception.

[0031] When performing reverse filtering on the system calibration data, the control module traverses the set of singular values ​​obtained in the previous steps for a specific wavelength point, such as 1550nm, in memory. Since the 2×2 Jones matrix contains at most two non-zero singular values, the maximum value is extracted through comparison logic and designated as the maximum singular value index. This maximum singular value represents the electric field amplitude transmission coefficient of the main polarization state of the test system at that wavelength, and its square characterizes the optical power transmission capability of the corresponding main polarization state. Assuming the maximum value obtained is 0.98, to filter out low-order small-signal interference caused by thermal fluctuations, fiber perturbations, and instrument noise floor in the measurement optical path, a scaling percentage factor is input, either set by the user or calculated by the system based on the signal-to-noise ratio. In applications testing the isolation of optical isolators, this factor can be set between 0.5% and 5%, with 1.5% being an example. The regularization threshold is calculated by multiplying the maximum singular value index by the coefficient. The threshold is calculated to be 0.0147 according to the above values. This threshold is configured as the lower limit interception benchmark in the truncated singular value decomposition inversion algorithm. Any singular value component smaller than this value will be regarded as a noise component and set to zero, so as to avoid the high-frequency noise amplification effect caused by ill-conditioned matrix inversion and maintain the numerical stability of the solution.

[0032] S3. The matrix elements of the forward and reverse Jones matrices are used as complex-valued functions of wavelength and their magnitudes and phase angles are separated. A complex model function is constructed using a non-negative amplitude mode function containing a Fabry-Perot interference periodic term and a Cauchy dispersion slowly varying phase angle term. The function is fitted with a weighted attenuation factor that decreases as the relative energy of discarded singular values ​​increases, thereby obtaining a continuous function expression of the matrix elements as a function of wavelength. Based on the continuous function expression, the isolation wavelength characteristic curve of the optical isolator within a preset wavelength range is calculated.

[0033] Four matrix data sequences are extracted from the forward and reverse Jones matrices, respectively. For each matrix data sequence, the modulus of the matrix element is calculated, and the phase angle of the matrix element is also calculated. The phase angle is then unwrapped and normalized relative to the relative phase using a preset reference matrix element or reference wavelength point as a reference. A complex model function is constructed. The modulus of this function is set as a non-negative electric field amplitude transmission coefficient function composed of a slowly varying amplitude envelope term and a Fabry-Perot interference periodic ripple term. The Fabry-Perot interference periodic ripple term can be represented in cosine ripple form, or the corresponding electric field amplitude modulus can be obtained by taking the square root of the Fabry-Perot Airy power transmittance. The overall amplitude modulus function is ensured to be non-zero through amplitude limiting, square root, or positive parameterization to ensure that the matrix element modulus represents the electric field amplitude transmission coefficient rather than the power transmittance. The phase angle of this function is set as a slowly varying Cauchy dispersion term, with the form: Alternatively, an equivalent expression for the relative phase angle can be used, where... Let A represent the Cauchy dispersion refractive index, A be the dimensionless zeroth-order Cauchy dispersion coefficient, and B and C be the second-order and fourth-order Cauchy dispersion coefficients, respectively. Indicates the effective optical path difference. Let λ represent the constant phase term, and λ is related to... Before being substituted into the calculation, all values ​​are uniformly converted to the same length unit. Those skilled in the art should know that, before being substituted into the calculation, the wavelength λ and effective optical path difference are uniformly converted to length units matching the dispersion coefficients, based on the dimensions defined by the Cauchy dispersion coefficients B and C. The sum of squares of the singular values ​​set to zero when constructing the pseudo-inverse matrix is ​​calculated, and the ratio of this sum to the sum of squares of all singular values ​​is calculated as the relative energy index for discarding singular values. The relative energy index is input into a decreasing weighting function, for example... ,in To discard singular values ​​relative energy indices, α is a positive adjustment coefficient and ε is a regularization constant, ensuring that ill-conditioned matrices or wavelengths with high energy receive smaller weights. A least-squares fitting function is invoked, substituting the weight parameters into the loss function to optimize the matrix element modulus and phase angle data, iteratively outputting the optimal parameters for the non-negative amplitude modulus function and the Cauchy dispersion slowly varying phase angle term. These optimal parameters are then substituted into the complex model function to obtain the continuous function expression of each matrix element as a function of wavelength. Based on the continuous function expressions of the forward and reverse Jones matrices, the transmittance parameters are calculated, and the ratio of the forward transmittance parameter to the reverse transmittance parameter is calculated. A base-10 logarithmic function is used to take the logarithm of this ratio, and then multiplied by 10 to obtain the continuous isolation wavelength characteristic curve of the optical isolator within a preset wavelength range. In an optional embodiment, the step of using the matrix elements of the positive and negative Jones matrices as complex-valued functions of wavelength and separating their magnitudes and phase angles, and constructing a complex model function using a non-negative amplitude magnitude function containing Fabry-Perot interferometry periodic terms and a Cauchy dispersion slowly varying phase angle term, includes: Based on the Cauchy dispersion relation and the relationship between the propagation optical path and wavelength, an expression for the slowly varying Cauchy dispersion term is constructed. Based on the periodic spectral ripple caused by multiple reflection interference, an expression for the periodic ripple term of the Fabry-Perot interference is constructed. The reference amplitude term, the slowly varying amplitude envelope term, and the Fabry-Perot interference periodic ripple term are combined to generate a non-negative amplitude mode expression. Using the Cauchy dispersion slow variable expression as the phase angle, construct a complex exponential term; Multiplying the nonnegative amplitude modulus expression by the complex exponential term yields the corresponding complex model function.

[0034] A parameterized mathematical model can be established for the transmission characteristics of the core magneto-optical crystal and the front and rear polarizers of the optical isolator. For the phase delay characteristics, the second-order Cauchy dispersion formula is used, which includes material dispersion coefficients. For example, the first coefficient for yttrium iron garnet crystal in the near-infrared band is approximately 2.2, and the second coefficient is approximately 0.03. This dispersion model is combined with the crystal length of 2 mm and the wavenumber term corresponding to 2π / λ is introduced. Before calculation, the crystal length and wavelength are converted to the same unit of length to construct an expression for the slowly varying Cauchy dispersion term representing the transmission phase delay. This formula describes the nonlinear variation of the phase angle with wavelength. Considering the multi-beam interference effect generated by the multi-layer coating interface and crystal end face inside the isolator, a cosine periodic oscillation equation including interference contrast parameters and effective optical path difference parameters of the interference cavity is used to construct an expression for the Fabry-Perot interference periodic ripple term. The reference amplitude transmittance coefficient is approximately 0.95, and the parameter range representing the reflection crosstalk intensity can be set from 0.001 to 0.05. The effective optical path difference parameter of the interference cavity refers to the equivalent optical path difference between the propagation paths of adjacent reflected beams in multi-beam interference, used to determine the period of the Fabry-Perot interference ripple as a function of wavelength. The reference amplitude term, the slowly varying amplitude envelope term, and the Fabry-Perot interference periodic ripple term are combined to form a non-negative amplitude mode expression. The Cauchy dispersion slowly varying term is set as the phase angle, and these terms are combined into a complex model function through complex multiplication. This composite model covers the envelope attenuation and ripple oscillation characteristics of the isolator matrix elements over a wide wavelength range. The Fabry-Perot periodic ripple component curves are shown below. Figure 3 As shown.

[0035] In an optional embodiment, the step of fitting the matrix elements using a decay factor that decreases as the relative energy of discarded singular values ​​increases, as weights, to obtain a continuous functional expression of the matrix elements as a function of wavelength, includes: Obtain the matrix element modulus and phase angle observation values ​​of the forward and reverse Jones matrices at each test wavelength point; The data fitting weights corresponding to each test wavelength point are determined based on the attenuation factor. Based on the data fitting weights, a weighted residual objective function is constructed between the predicted values ​​of the complex model function and the observed values. The model parameters of the complex model function are obtained by minimizing the weighted residual objective function. Substituting the model parameters into the complex model function generates a continuous function expression of the matrix elements as a function of wavelength.

[0036] During the numerical optimization process of parameter fitting, the modulus and phase angle experimental data obtained from the decomposition of the Jones matrix elements in the previous discrete tests, for example, 501 data points in the range of 1520nm to 1570nm, are imported into the iterative analysis module. The phase angle experimental data are relative phase angle data after phase unwrapping, excluding the global absolute optical phase that cannot be uniquely determined by Stokes intensity measurements. For the noise level differences at different wavelengths, an attenuation factor associated with the truncated singular value attenuation mechanism is input as a data confidence weight; specifically, the sum of squares of the truncated singular values ​​can be calculated first. Then calculate the sum of squares of all singular values. ,Will As a relative energy index that discards singular values, and weights are set... Therefore, wavelengths with ill-conditioned matrices or high proportions of truncated energy are assigned lower weights (less than 1), while regions with good signal quality are assigned weights close to 1.0. In the iterative calculation phase, the objective function is set as minimizing the weighted sum of squared residuals. Reasonable initial values ​​are assigned to each parameter, and an optimization iteration combining gradient descent and the Gauss-Newton method is performed until the L2 norm of the residual gradient is less than the set value. The convergence threshold is reached or the maximum number of iterations is reached (1000). The converged feature coefficient vector is extracted. For example, if the calculated dispersion coefficient values ​​are 2.185 and 0.029, this feature coefficient vector is substituted back into the complex model function to generate a continuous function expression for the optical isolator matrix elements that spans the entire preset wavelength band, is not limited by the measurement step size, and controls random noise as a function of wavelength.

[0037] In an optional embodiment, calculating the isolation wavelength characteristic curve of the optical isolator within a preset wavelength range based on the continuous function expression includes: Substituting the target wavelength into the continuous function expressions of the forward and reverse Jones matrices, we obtain the forward and reverse Jones matrices for the corresponding wavelength point. Under unpolarized light conditions, the forward and reverse light power transmittances are calculated based on half the trace of the product of the Jones matrix and its conjugate transpose. The ratio of the forward light power transmittance to the reverse light power transmittance is used as the transmittance. The transmittance is calculated using a logarithmic transformation formula to obtain the isolation value at the corresponding wavelength point; Based on each wavelength point and its corresponding isolation value, an isolation wavelength characteristic curve of the optical isolator within a preset wavelength range is generated.

[0038] During the performance evaluation of the optical isolator, the processing system can generate a continuous performance spectrum according to a preset wavelength resolution, such as interpolation to a step size of 0.01 nm. The target wavelength to be calculated, such as 1550.00 nm, is substituted into the already determined continuous function expression to reconstruct the forward and reverse Jones matrices. To evaluate the average performance of the device in a real fiber optic communication network facing random polarization states, the forward optical power transmittance is calculated using a mathematical trace operation formula. This involves taking half the trace of the product of the forward Jones matrix and its conjugate transpose, and then using the same matrix operation to calculate the reverse optical power transmittance. For example, the calculated forward transmittance could be 0.891, while the reverse transmittance could be 0.000089. The transmittance ratio is obtained by performing a proportional operation on the optical power transmittance, and the isolation value representing the isolation capability is calculated using a logarithmic transformation formula. Continuing with the previous example, the isolation at this wavelength point is 40.0 dB. The algorithm iteratively calculates all interpolation nodes within a preset wavelength range, extracts the corresponding wavelength array as the x-axis and the isolation value array as the y-axis, and uses a visualization tool library to map and plot this coordinate data in a two-dimensional Cartesian coordinate system. This curve displays the peak isolation at the central operating wavelength and the roll-off trend on both sides, guiding the selection and evaluation of core passive components in optical communication systems.

[0039] In the second embodiment, the present invention also proposes a wavelength characteristic testing system for the isolation of an optical isolator, comprising the following modules: The measurement module is used to construct the optical path of the test system, which includes a tunable light source, a polarization state generator, and a polarization state analyzer. It acquires the test light output by the tunable light source within a preset wavelength range. When no optical isolator under test is set in the optical path of the test system, it measures the system calibration Jones matrix at each wavelength point based on the input and output reference plane of the optical isolator under test, and performs singular value decomposition to obtain the singular values ​​at each wavelength point. The optical isolator under test is placed into the optical path of the test system along the forward and reverse transmission directions, respectively, and the forward mixed Jones matrix and the reverse mixed Jones matrix at each wavelength point are measured and obtained. The construction module is used to set a regularization threshold based on the maximum singular value of the system calibration Jones matrix at each wavelength point, discard singular values ​​smaller than the threshold to construct a pseudo-inverse matrix of the calibration Jones matrix, and multiply the pseudo-inverse matrix with the corresponding hybrid Jones matrix to calculate the forward Jones matrix and the reverse Jones matrix of the optical isolator under test at the wavelength point. The calculation module is used to treat the matrix elements of the forward and reverse Jones matrices as complex-valued functions of wavelength and to separate the magnitude and phase angle. It constructs a complex model function using a non-negative amplitude magnitude function containing a Fabry-Perot interference periodic term and a Cauchy dispersion slowly varying phase angle term, and fits the function with a decay factor that decreases as the relative energy of discarded singular values ​​increases, to obtain a continuous function expression of the matrix elements changing with wavelength. Based on the continuous function expression, the isolation wavelength characteristic curve of the optical isolator within a preset wavelength range is calculated.

[0040] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for testing the wavelength characteristics of isolation of an optical isolator, characterized in that, Includes the following steps: A test system optical path is constructed, including a tunable light source, a polarization state generator, and a polarization state analyzer. The test light output by the tunable light source within a preset wavelength range is obtained. When the optical isolator under test is not set in the test system optical path, the system calibration Jones matrix of each wavelength point is measured based on the input and output reference plane of the optical isolator under test, and singular value decomposition is performed to obtain the singular values ​​of each wavelength point. The optical isolator under test is placed into the test system optical path along the forward and reverse transmission directions, respectively, and the forward mixed Jones matrix and the reverse mixed Jones matrix of each wavelength point are measured. For each wavelength point, a regularization threshold is set based on the maximum singular value of the system calibration Jones matrix at the wavelength point. Singular values ​​smaller than the threshold are discarded to construct a pseudo-inverse matrix of the calibration Jones matrix. The pseudo-inverse matrix is ​​multiplied by the corresponding hybrid Jones matrix to calculate the forward Jones matrix and the reverse Jones matrix of the optical isolator under test at the wavelength point. The matrix elements of the forward and reverse Jones matrices are used as complex-valued functions of wavelength, and their magnitudes and phase angles are separated. A complex model function is constructed using a non-negative amplitude mode function containing a Fabry-Perot interference periodic term and a Cauchy dispersion slowly varying phase angle term. The function is fitted with a weight that decreases as the relative energy of discarded singular values ​​increases, thus obtaining a continuous function expression of the matrix elements as a function of wavelength. Based on the continuous function expression, the isolation wavelength characteristic curve of the optical isolator within a preset wavelength range is calculated.

2. The method according to claim 1, characterized in that, The acquisition of test light output from a tunable light source within a preset wavelength range includes: Set the starting and ending test wavelengths of the tunable light source as the lower and upper limits of measurement, respectively. Set the step interval and tuning scan speed for wavelength scanning; Wavelength scanning control commands are sent to the tunable light source via the communication interface; The tunable light source is controlled to be tuned point by point between the lower and upper measurement limits according to the stated step interval; A continuous monochromatic test light is output at each wavelength dwell point and injected into the optical path of the test system.

3. The method according to claim 1, characterized in that, When no optical isolator under test is set in the optical path of the test system, the system calibration Jones matrix at each wavelength point is measured based on the input and output reference plane of the optical isolator under test, and singular value decomposition is performed to obtain the singular values ​​at each wavelength point, including: Use an unloaded fiber optic patch cord to connect the polarization state generator and the polarization state analyzer to form a blank test optical path without the device under test; Through input-side pre-compensation, the input polarization state is equivalently defined on the input reference plane of the optical isolator under test; At each test wavelength point, the polarization state generator is controlled to output horizontally linearly polarized light, vertically linearly polarized light, 45° linearly polarized light, and right-hand circularly polarized light in sequence. The output Stokes vector data corresponding to each input polarization state is acquired using a polarization state analyzer. Based on the output Stokes vector data, the system calibration Jones matrix for the corresponding wavelength point is obtained using the Jones matrix derivation algorithm. The system calibration Jones matrix is ​​subjected to singular value decomposition, and the resulting singular values ​​are extracted as singular value parameters for the wavelength points.

4. The method according to claim 3, characterized in that, The step of setting a regularization threshold for each wavelength point based on the maximum singular value of the system calibration Jones matrix at that wavelength point includes: Iterate through all non-zero singular values ​​obtained from the singular value decomposition of the system calibration Jones matrix at each wavelength point, extract the singular value with the largest value as the maximum singular value index, multiply the maximum singular value index by a preset scaling percentage coefficient, and use the resulting product value as the regularization threshold for lower limit interception.

5. The method according to claim 1, characterized in that, The method of using the matrix elements of the positive and negative Jones matrices as complex-valued functions of wavelength and separating their magnitudes and phase angles, and constructing a complex model function using a non-negative amplitude magnitude function containing Fabry-Perot interferometry periodic terms and a Cauchy dispersion slowly varying phase angle term, includes: Based on the Cauchy dispersion relation and the relationship between the propagation optical path and wavelength, an expression for the slowly varying Cauchy dispersion term is constructed. Based on the periodic spectral ripple caused by multiple reflection interference, an expression for the periodic ripple term of the Fabry-Perot interference is constructed. The reference amplitude term, the slowly varying amplitude envelope term, and the Fabry-Perot interference periodic ripple term are combined to generate a non-negative amplitude mode expression. Using the Cauchy dispersion slow variable expression as the phase angle, construct a complex exponential term; Multiplying the nonnegative amplitude modulus expression by the complex exponential term yields the corresponding complex model function.

6. The method according to claim 1, characterized in that, The process of fitting a continuous function expression of matrix elements as a function of wavelength, using a decay factor that decreases with increasing relative energy of discarded singular values ​​as weights, includes: Obtain the matrix element modulus and phase angle observation values ​​of the forward and reverse Jones matrices at each test wavelength point; The data fitting weights corresponding to each test wavelength point are determined based on the attenuation factor. Based on the data fitting weights, a weighted residual objective function is constructed between the predicted values ​​of the complex model function and the observed values. The model parameters of the complex model function are obtained by minimizing the weighted residual objective function. Substituting the model parameters into the complex model function generates a continuous function expression of the matrix elements as a function of wavelength.

7. The method according to claim 1, characterized in that, The step of calculating the isolation wavelength characteristic curve of the optical isolator within a preset wavelength range based on the continuous function expression includes: Substituting the target wavelength into the continuous function expressions of the forward and reverse Jones matrices, we obtain the forward and reverse Jones matrices for the corresponding wavelength point. Under unpolarized light conditions, the forward and reverse light power transmittances are calculated based on half the trace of the product of the Jones matrix and its conjugate transpose. The ratio of the forward light power transmittance to the reverse light power transmittance is used as the transmittance. The transmittance is calculated using a logarithmic transformation formula to obtain the isolation value at the corresponding wavelength point; Based on each wavelength point and its corresponding isolation value, an isolation wavelength characteristic curve of the optical isolator within a preset wavelength range is generated.

8. A system for testing the wavelength characteristics of the isolation of an optical isolator, characterized in that, Includes the following modules: The measurement module is used to construct the optical path of the test system, which includes a tunable light source, a polarization state generator, and a polarization state analyzer. It acquires the test light output by the tunable light source within a preset wavelength range. When no optical isolator under test is set in the optical path of the test system, it measures the system calibration Jones matrix at each wavelength point based on the input and output reference plane of the optical isolator under test, and performs singular value decomposition to obtain the singular values ​​at each wavelength point. The optical isolator under test is placed into the optical path of the test system along the forward and reverse transmission directions, respectively, and the forward mixed Jones matrix and the reverse mixed Jones matrix at each wavelength point are measured and obtained. The construction module is used to set a regularization threshold based on the maximum singular value of the system calibration Jones matrix at each wavelength point, discard singular values ​​smaller than the threshold to construct a pseudo-inverse matrix of the calibration Jones matrix, and multiply the pseudo-inverse matrix with the corresponding hybrid Jones matrix to calculate the forward Jones matrix and the reverse Jones matrix of the optical isolator under test at the wavelength point. The calculation module is used to treat the matrix elements of the forward and reverse Jones matrices as complex-valued functions of wavelength and to separate the magnitude and phase angle. It constructs a complex model function using a non-negative amplitude magnitude function containing a Fabry-Perot interference periodic term and a Cauchy dispersion slowly varying phase angle term, and fits the function with a decay factor that decreases as the relative energy of discarded singular values ​​increases, to obtain a continuous function expression of the matrix elements changing with wavelength. Based on the continuous function expression, the isolation wavelength characteristic curve of the optical isolator within a preset wavelength range is calculated.

9. The system according to claim 8, characterized in that, The acquisition of test light output from a tunable light source within a preset wavelength range includes: Set the starting and ending test wavelengths of the tunable light source as the lower and upper limits of measurement, respectively. Set the step interval and tuning scan speed for wavelength scanning; Wavelength scanning control commands are sent to the tunable light source via the communication interface; The tunable light source is controlled to be tuned point by point between the lower and upper measurement limits according to the stated step interval; A continuous monochromatic test light is output at each wavelength dwell point and injected into the optical path of the test system.

10. The system according to claim 8, characterized in that, When no optical isolator under test is set in the optical path of the test system, the system calibration Jones matrix at each wavelength point is measured based on the input and output reference plane of the optical isolator under test, and singular value decomposition is performed to obtain the singular values ​​at each wavelength point, including: Use an unloaded fiber optic patch cord to connect the polarization state generator and the polarization state analyzer to form a blank test optical path without the device under test; Through input-side pre-compensation, the input polarization state is equivalently defined on the input reference plane of the optical isolator under test; At each test wavelength point, the polarization state generator is controlled to output horizontally linearly polarized light, vertically linearly polarized light, 45° linearly polarized light, and right-hand circularly polarized light in sequence. The output Stokes vector data corresponding to each input polarization state is acquired using a polarization state analyzer. Based on the output Stokes vector data, the system calibration Jones matrix for the corresponding wavelength point is obtained using the Jones matrix derivation algorithm. The system calibration Jones matrix is ​​subjected to singular value decomposition, and the resulting singular values ​​are extracted as singular value parameters for the wavelength points.

Images