Network damage identification method, device and equipment

Abstract

This application provides a method, apparatus, and device for network impairment identification, applicable to the field of communications. The method includes: receiving multiple signals transmitted by a transmitter through multiple transmitting ports, wherein each signal corresponds one-to-one with a transmitting port, and one of the signals is received by at least one of the receiving ports; compensating the multiple signals using a compensation model to obtain compensation results; updating the parameters of the compensation model in a way that improves the compensation results; acquiring the parameters of the compensation model when the compensation results meet preset conditions; and identifying network system impairments based on the acquired parameters of the compensation model. This method can identify network impairments efficiently and at low cost.

Description

Technical Field

[0001] This application relates to the field of communications, and in particular to a method, apparatus, and device for identifying network impairments. Background Technology

[0002] With the development of network technology, symbol periods are constantly shortening and modulation formats are continuously evolving, making the impact of network impairments on network communication performance increasingly significant. To maintain network communication performance, it is necessary to promptly repair network impairments, which requires timely identification of the location of these impairments.

[0003] Current solutions for identifying network impairments either require the introduction of additional hardware, which is costly and can cause communication interruptions; or they are opaque to the signal format, requiring a full understanding of the signal format, making them difficult to apply in scenarios where the signal format is unknown or changes rapidly. Summary of the Invention

[0004] This application provides a method, apparatus, and device for identifying network impairments, which can identify network impairments efficiently and at low cost.

[0005] Firstly, a network impairment identification method is provided, applied to a receiver in a network system. The network system also includes a transmitter. The receiver has multiple receiver ports, and the transmitter has multiple transmitter ports, wherein the number of receiver ports is greater than or equal to the number of transmitter ports. The method includes: receiving multiple signals transmitted by the transmitter through the multiple transmitter ports via the multiple receiver ports, wherein each signal corresponds one-to-one with a transmitter port, and one of the multiple signals is received by at least one of the multiple receiver ports; compensating the multiple signals using a compensation model to obtain compensation results for the multiple signals; updating the parameters of the compensation model in a direction that improves the compensation results; obtaining the parameters of the compensation model when the compensation results meet preset conditions; and identifying impairments in the network system based on the obtained parameters of the compensation model.

[0006] The damage identified by this method can refer to damage at the physical layer, i.e., damage to the hardware.

[0007] Impairment in a network system causes multiple signals to affect each other, and these effects are reflected in the parameters of a compensation model that can compensate the signals of these multiple signals to meet preset conditions. This method estimates or explores the parameters of the compensation model that can compensate the signals of multiple signals to meet preset conditions, and based on these parameters, identifies network system impairments. This achieves network impairment identification without requiring additional hardware in the network system or knowledge of the signal format. This method is low-cost, the network impairment identification method itself has no impact on communication performance, and it is format-transparent, making it widely applicable.

[0008] In one possible implementation, the condition includes: the compensation result is greater than or equal to a preset threshold, or the compensation result reaches the optimal result that can be achieved by using the compensation model to compensate multiple signals.

[0009] In one possible implementation, identifying damage to the network system based on the parameters of the acquired compensation model includes: confirming that damage has occurred in the network system based on the parameters of the acquired compensation model.

[0010] Whether a network system is damaged or not causes multiple signals to have different effects on each other. These effects are reflected in the parameters of a compensation model that can compensate the signals of multiple signals to meet preset conditions. Therefore, the presence or absence of damage to the network system can be identified based on the parameters of the compensation model that compensates the signals of multiple signals to meet preset conditions.

[0011] In one possible implementation, identifying impairments in the network system based on the parameters of the acquired compensation model includes: identifying the location of the impairment in the network system, including the receiver and / or transmitter, based on the parameters of the acquired compensation model; and / or identifying the degree of impairment based on the parameters of the acquired compensation model; and / or identifying the type of impairment based on the parameters of the acquired compensation model; wherein the type of impairment includes polarization imbalance impairment.

[0012] Different locations, degrees, and types of damage in a network system cause multiple signals to have different effects on each other. These effects are reflected in the parameters of a compensation model that can compensate the signals of multiple signals to meet preset conditions. Therefore, based on the parameters of the compensation model that compensates the signals of multiple signals to meet preset conditions, the location, degree, and type of damage in the network system can be identified.

[0013] In one possible implementation, crosstalk occurs between multiple signals during transmission; compensation for the multiple signals includes increasing the non-Gaussian properties of the multiple signals. The compensation model used for compensating the multiple signals can be obtained based on the Independent Component Analysis (ICA) algorithm.

[0014] In this implementation, the polarization transmission effect inherent in MIMO systems can be utilized to increase the non-Gaussian properties of the multiple signals, thereby achieving signal compensation for the multiple signals. This allows for the estimation or exploration of the parameters of a compensation model that can compensate the multiple signals to the preset conditions, and based on these parameters, damage to the network system can be identified.

[0015] In one possible implementation, updating the parameters of the compensation model in order to improve the compensation results includes updating the parameters of the compensation model with the aim of maximizing the non-Gaussian properties of the multiple signals.

[0016] In this implementation, by maximizing the non-Gaussian properties of the multiple signals to update the parameters of the compensation model, the parameters of the compensation model that can compensate the multiple signals to meet the preset conditions can be obtained.

[0017] In one possible implementation, the compensation model includes multiple tap matrices; using the compensation model to compensate multiple signals and obtain the compensation results of the multiple signals includes: constructing a vector matrix of the multiple signals; using multiple tap matrices to calculate the vector matrix of the multiple signals and obtain the signal compensation results of the multiple signals.

[0018] The signal compensation result is characterized as follows:

[0019]

[0020] Among them, S OUT This represents the signal compensation result, where x represents the first signal in the multi-channel signal set, y represents the second signal in the multi-channel signal set, M represents the tap matrix, the superscript * represents the conjugate operation, and the superscript T represents the transpose operation. The vector matrix represents the signals in the multiple signals, and n represents the time index of the signal in the multiple signals;

[0021] The tap matrix is ​​represented as follows:

[0022]

[0023] Where M′ represents a matrix composed of multiple tap matrices, H represents the parameter in the tap matrix used to represent crosstalk between signals, L is the difference obtained by subtracting 1 from the length of the input data required by the tap matrix, L is an integer greater than 1, and L is an even number.

[0024] In this implementation, the tap matrix in the compensation model also functions as a filter, performing equalization and demultiplexing on the input signal. Multiple tap matrices comprising the compensation model form a butterfly filter result, where the tap coefficients of each filter are pairwise correlated with transceiver impairments, thus aiding in transceiver impairment separation. Here, transceiver refers to the receiving and transmitting ends, and transceiver impairments refer to both receiving and transmitting impairments. Furthermore, the butterfly filter algorithm architecture based on ICA (Inter-Cyclic Algorithm) is beneficial for utilizing the polarization crosstalk naturally present in network systems to achieve impairment decomposition and monitoring, while simultaneously resolving the amplitude ambiguity problem inherent in ICA.

[0025] In one possible implementation, constructing the vector matrix of the multiple signals includes: performing eigenvalue decomposition on the multiple signals to obtain a whitening matrix; using the whitening matrix to whiten the multiple signals to obtain whitened multiple signals; and based on the whitened multiple signals, constructing the vector matrix of the multiple signals using a time sliding window of L+1 units.

[0026] The whitening matrix is ​​represented as follows:

[0027]

[0028] Where V represents the whitening matrix, Λ represents the eigenvalue diagonal matrix obtained by eigenvalue decomposition of multiple signals, U represents the eigenvector matrix obtained by eigenvalue decomposition of multiple signals, the superscript T represents the transpose operation, and W represents the parameter in the whitening matrix used to represent crosstalk between signals.

[0029] The whitened multi-channel signals are characterized as follows:

[0030]

[0031] Among them, S IN This represents the signal from the multi-channel signal before whitening. This represents the signal in the whitened multi-channel signal;

[0032] The vector matrix representation of the multiple signals is as follows:

[0033]

[0034] in, The vector matrix represents the signals in the whitened multi-channel signal.

[0035] In this implementation, whitening processing ensures that the mean of different signals is 0 and the variance is 1, thereby achieving decorrelation between them. Here, correlation refers to a linear relationship. Decorrelation means eliminating any linear relationship between different signals. Furthermore, after whitening the multiple signals, a vector matrix of the multiple signals can be constructed using a time sliding window of L+1 units. The feature vector matrix constructed through this preprocessing scheme facilitates subsequent parameter updates, accelerates algorithm convergence, and ensures that the dimension of the input signal matches the subsequent tap matrix.

[0036] In one possible implementation, updating the parameters of the compensation model to improve the compensation result includes: normalizing the power of the signal compensation result using the first equation to obtain a normalized signal compensation result; and maximizing the non-Gaussian nature of the normalized signal compensation result using the second equation to update the parameters in the tap matrix.

[0037] The first form includes:

[0038]

[0039] in, This represents the normalized signal compensation result, and E represents the average operation.

[0040] The second form includes:

[0041]

[0042]

[0043]

[0044]

[0045] Where i∈[-L / 2 L / 2], This represents a nonlinear function. The superscript + indicates the updated parameters, and the superscript - indicates the parameters before the update.

[0046] In this implementation, the power of the signal compensation result is normalized by the first equation, and then the normalized signal compensation result is subjected to nonlinear operation by equation (7) to simulate the probability density distribution of the signal; then, the first and second derivatives of the nonlinear signal are calculated; then, the non-Gaussianity is maximized by Newton's method to update the tap coefficients of the tap matrix (i.e. the parameters of the compensation model), thereby estimating the parameters of the compensation model that can compensate the signals of multiple signals to meet the preset conditions.

[0047] In one possible implementation, identifying the location of network impairments in the network system based on the parameters of the acquired compensation model includes: obtaining a process matrix based on the parameters in the updated tap matrix; and identifying network impairments based on the process matrix using a third equation.

[0048] The process matrix is ​​represented as follows:

[0049]

[0050] The third form includes:

[0051]

[0052] τTx =(τ xx -τ yx +τ xy ) / 2

[0053] τ Rx =(τ xx -τ xy +τ yx ) / 2

[0054] g Tx =(E(|P xx |)E(|P xy |) / (E(|P yx |)E(|P yy |)))

[0055] g Rx =(E(|P xx |)E(|P yx |) / (E(|P xy |)E(|P yy |)))

[0056] Where P represents the process matrix corresponding to the tap matrix, P′ represents the matrix composed of process matrices corresponding to multiple tap matrices, T represents the parameters in the process matrix, τ represents the clock offset, g represents the amplitude imbalance, Tx represents the transmitting end, Rx represents the receiving end, f represents the frequency domain index, d represents the partial derivative, and f s The value represents the sampling rate, the superscript f indicates that the current variable belongs to the frequency domain, and FFT stands for Fourier transform.

[0057] In one possible implementation, identifying impairments in the network system based on the acquired parameters of the compensation model includes: identifying the location of the impairment in the network system based on the acquired parameters of the compensation model, the location including a receiver and / or a transmitter; the method further includes: sending the acquired parameters of the compensation model to the location, the acquired parameters of the compensation model being used to pre-compensate the signal at the location.

[0058] In this implementation, after identifying the location of the damage, the parameters of the compensation model that compensates multiple signals to meet preset conditions can be sent to the identified location. The acquired parameters of the compensation model are used to pre-compensate the signal at that location, thereby improving the ability to complete signal compensation in advance and to compensate the signal to meet preset conditions in advance, overcoming the damage to the signal caused by network damage and improving the communication performance of the network.

[0059] Secondly, a network impairment identification device is provided, configured at a receiver in a network system. The network system also includes a transmitter. The receiver has multiple receiver ports, and the transmitter has multiple transmitter ports. The number of receiver ports is greater than or equal to the number of transmitter ports in the transmitter. The device includes: a receiving unit for receiving multiple signals transmitted by the transmitter through the multiple transmitter ports, wherein each signal corresponds one-to-one with a transmitter port, and one of the multiple signals is received by at least one of the receiver ports; a compensation unit for compensating the multiple signals using a compensation model to obtain compensation results for the multiple signals; an update unit for updating the parameters of the compensation model in a direction that improves the compensation results; an acquisition unit for acquiring the parameters of the compensation model when the compensation results meet preset conditions; and an identification unit for identifying impairments in the network system based on the acquired parameters of the compensation model.

[0060] In one possible implementation, the condition includes: the compensation result is greater than or equal to a preset threshold, or the compensation result reaches the optimal result that can be achieved by using the compensation model to compensate multiple signals.

[0061] In one possible implementation, the identification unit is used to: confirm damage to the network system based on the parameters of the acquired compensation model.

[0062] In one possible implementation, the identification unit is used to: identify the location of the impairment in the network system based on the parameters of the acquired compensation model, the location including the receiver and / or transmitter; and / or, identify the degree of the impairment based on the parameters of the acquired compensation model; and / or, identify the type of impairment based on the parameters of the acquired compensation model; wherein the type of impairment includes polarization imbalance impairment.

[0063] In one possible implementation, crosstalk occurs between multiple signals during transmission; the compensation unit is used to increase the non-Gaussianity of the multiple signals.

[0064] In one possible implementation, the update unit is used to update the parameters of the compensation model with the aim of maximizing the non-Gaussian properties of the multiple signals.

[0065] Thirdly, a communication device is provided, comprising: a memory and a processor; the memory for storing computer instructions; and the processor for executing the computer instructions stored in the memory to implement the method provided in the first aspect.

[0066] Fourthly, a chip is provided for performing the method provided in the first aspect.

[0067] Fifthly, a computer storage medium is provided, the computer storage medium including computer instructions, which, when executed on a communication device, cause the communication device to perform the method provided in the first aspect.

[0068] In a sixth aspect, a computer program product is provided, wherein the program code contained in the computer program product, when executed by a processor in a communication device, implements the method provided in the first aspect.

[0069] The beneficial effects of aspects two through six can be found in the above description of the beneficial effects of aspect one, and will not be repeated here. Attached Figure Description

[0070] Figure 1 A schematic diagram of the structure of a network system provided in an embodiment of this application;

[0071] Figure 2 A schematic diagram of the structure of a network system provided in an embodiment of this application;

[0072] Figure 3 A flowchart illustrating a network impairment identification method provided in this application embodiment;

[0073] Figure 4 A schematic diagram illustrating a network impairment identification method provided in an embodiment of this application;

[0074] Figure 5 This is a schematic diagram of the structure of a network damage identification device provided in an embodiment of this application;

[0075] Figure 6 This is a schematic diagram of the structure of a communication device provided in an embodiment of this application. Detailed Implementation

[0076] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of them. In the embodiments of this application, "multiple" refers to two or more.

[0077] This application provides a network impairment identification method applicable to multiple-input multiple-output (MIMO) network systems. This network system can transmit multiple signals simultaneously on the same or multiple channels. When the receiving end of the network system receives these multiple signals, it can compensate for them using a preset compensation model to obtain the compensation result. The parameters of the compensation model are updated in the direction of improving the compensation result. The updated compensation model is then used to compensate for the multiple signals again. This process is repeated until the compensation result of the multiple signals meets a preset condition. At this point, the parameters of the compensation model that make the compensation result of the multiple signals meet the preset condition are obtained, and based on the obtained parameters, it is determined whether the network system has suffered impairment, and the location, degree, and type of impairment, thus realizing network system impairment identification. Here, signal compensation refers to impairment compensation of the signal. Specifically, the signal received by the receiving end may be an impaired signal. Compensating for the impairment, or making up for the impairment, to improve or enhance the signal quality can be called impairment compensation.

[0078] It is understandable that whether a network system is damaged, and the different locations, degrees, and types of damage, will cause multiple signals to have different effects on each other (e.g., different degrees of crosstalk). These effects are reflected in the parameters of a compensation model that can compensate the multiple signals to meet preset conditions. Thus, by using the parameters of the compensation model that makes the compensation results of the multiple signals meet preset conditions, damage to the network system can be identified.

[0079] Next, the network impairment identification method provided in the embodiments of this application will be described in detail.

[0080] Figure 1 An embodiment of the network system provided in this application is illustrated. For example... Figure 1 As shown, the network system includes a transmitter 110 and a receiver 120. The transmitter can also be called a transceiver (Tx), and the receiver can also be called a receiver (Rx).

[0081] The transmitting end 110 includes a processor 111 and multiple transmitting ports, such as transmitting port 112 and transmitting port 113. The receiving end 120 includes a processor 121 and multiple receiving ports, such as receiving port 121 and receiving port 122. The number of receiving ports in the receiving end 120 is greater than or equal to the number of transmitting ports in the transmitting end 110.

[0082] In the following description, when there is no special distinction between transmitting port 112 and transmitting port 113, they may be simply referred to as transmitting ports. When there is no special distinction between receiving port 121 and receiving port 122, they may be simply referred to as receiving ports.

[0083] Processor 111 is used to transmit multiple signals through multiple transmit ports, with each transmit port capable of transmitting one signal. The transmit ports can convert digital signals from processor 111 into analog signals and transmit them.

[0084] In the embodiments of this application, one signal can be understood as one signal.

[0085] The processor 121 can receive multiple signals transmitted by the transmitter 110 through multiple receiving ports. Among these multiple signals, one signal is received by at least one receiving port, that is, one signal can be received by one receiving port or by multiple receiving ports.

[0086] The receiving port can convert the received analog signal into a digital signal and transmit the digital signal to the processor 121. The processor 121 can process the digital signal and use it to identify damage to the network system. Details will be provided below.

[0087] In some embodiments, Figure 1 The network system shown can be an optical coherent communication system using a MIMO architecture, that is, the network impairment identification method provided in this application embodiment can be applied to an optical coherent communication system using a MIMO architecture.

[0088] Figure 2 This paper illustrates a specific implementation of an optical coherent communication system employing a MIMO architecture. For example, the processor 111 at the transmitting end can be implemented as a digital signal processing (DSP), referred to as a Tx-DSP, and the processor 121 at the receiving end can also be implemented as a DSP, referred to as an Rx-DSP.

[0089] In one example, such as Figure 2As shown, the processor 111 may include a random signal module, a pulse shaping module, a pre-emphasis module, and a transmitter impairment pre-comp module. The random signal module generates the signal to be transmitted, the pulse shaping module shapes the signal, the pre-emphasis module pre-emphasizes the signal, and the transmitter impairment pre-comp module pre-computes the signal. After these processes, a digital signal to be transmitted is obtained. The processor 111 can generate multiple signals simultaneously, and each signal can be transmitted through a single transmission port.

[0090] like Figure 2 As shown, each of the multiple transmission ports of the transmitter 110 includes a digital-to-analog converter (DAC) and an electrical amplifier. The DAC converts the digital signal corresponding to that transmission port into an analog signal, and the electrical amplifier amplifies the analog signal.

[0091] like Figure 2 As shown, the transmitter 110 also includes a dual-polarization-IQ modulator (DP-IQ modulator). Signals to be transmitted from multiple transmitter ports are modulated by the dual-polarization IQ modulator to obtain modulated optical signals. The modulated optical signals from different paths are orthogonal.

[0092] The modulated optical signal is sent into the optical fiber link and arrives at the receiving end 120 after passing through several amplifiers, optical fibers and filters.

[0093] like Figure 2 As shown, the receiver 120 includes an integrated coherent receiver (ICR). The integrated coherent receiver receives optical signals and uses the local oscillator light and the received optical signals to perform beat frequency conversion and photoelectric conversion.

[0094] Each of the multiple receiving ports in receiver 120 includes an analog-to-digital converter (ADC). The ADC of each receiving port converts the analog signal (the analog signal output by the ICR) corresponding to that receiving port into a digital signal, thus enabling signal reception. The receiving port then sends the received digital signal to processor 121 for processing.

[0095] like Figure 2As shown, the processor 121 includes, in sequence according to the signal transmission direction, a receiver impairment pre-component (Rximba.pre-comp) module, a resampling module, a chromatic dispersion compensation (CDC) module, and an analysis module. In one example, the analysis module is specifically an independent component analysis (ICA) module.

[0096] In this process, after the signal passes through the receiver impairment pre-compensation module, resampling module, and dispersion compensation module, the analysis module identifies network impairments according to the network impairment identification method provided in the embodiments of this application.

[0097] In some embodiments, the network system can adopt other communication systems with MIMO architecture. That is, the network impairment identification method provided in this application embodiment can be applied to other communication systems with MIMO architecture, such as mode division multiplexing optical fiber communication systems, wireless communication systems, speech signal processing systems, and biomedical signal processing systems.

[0098] The above example illustrates a network system that can be used to implement the network identification method provided in the embodiments of this application. Next, the flow of the network identification method provided in the embodiments of this application will be described in detail.

[0099] This method can be applied to the receiving end of a network system; that is, the method is executed by the receiving end of the network system, for example... Figure 1 The receiver 120 shown is executed. The system network system also includes a transmitter, wherein the receiver has multiple receiver ports and the transmitter has multiple transmitter ports, and the number of receiver ports is greater than or equal to the number of transmitter ports in the transmitter.

[0100] like Figure 3 As shown, the method includes the following steps: Step 301, receiving multiple signals transmitted by the transmitter through the multiple receiving ports, wherein each of the multiple signals corresponds one-to-one with the multiple transmitting ports, and one of the multiple signals is received by at least one of the multiple receiving ports; Step 302, compensating the multiple signals using a compensation model to obtain a compensation result for the multiple signals; Step 303, updating the parameters of the compensation model in a direction that improves the compensation result; Step 304, obtaining the parameters of the compensation model when the compensation result meets a preset condition; Step 305, identifying damage to the network system based on the obtained parameters of the compensation model.

[0101] Next, we will describe each of the above steps in detail with reference to different embodiments.

[0102] First, in step 301, multiple signals transmitted by the transmitting end through the multiple receiving ports are received through the multiple receiving ports, wherein each of the multiple signals corresponds one-to-one with the multiple transmitting ports, and one of the multiple signals is received by at least one of the multiple receiving ports.

[0103] In a network system, a transmitting port at the transmitting end can transmit one signal, which can be received by at least one receiving port at the receiving end. Furthermore, one receiving port receives one signal. That is, a signal can be received by one receiving port, or by two or more receiving ports. For example, in unicast, one signal is received by one receiving port, while in broadcast, one signal is received by two or more receiving ports.

[0104] The different signals in this multi-channel signaling system are orthogonal. A single signal refers to multiple signals transmitted sequentially in the time domain.

[0105] In some embodiments, the network system can be a 2×2 MIMO system, meaning it has two transmit ports and two receive ports. The signal transmitted by one of the transmit ports can be called the X signal, and the signal transmitted by the other transmit port can be called the Y signal; that is, the X signal is one path, and the Y signal is another. One of the receive ports receives the X signal, and the other receives the Y signal. The polarization directions of the X signal and the Y signal are orthogonal.

[0106] In some embodiments, the multiple signals can be transmitted over the same physical link. Taking optical communication as an example, the multiple signals are transmitted concurrently in the same optical fiber.

[0107] Next, 302, the compensation model is used to compensate the multiple signals to obtain the compensation result of the multiple signals.

[0108] As mentioned above, signal compensation can refer to damage compensation of the signal. Specifically, the signal received by the receiving end may be a damaged signal. Compensating for the damage of the signal or making up for the damage of the signal to improve or enhance the quality of the signal can be called damage compensation of the signal.

[0109] Compensating multiple signals refers to simultaneously compensating each individual signal within the multiple signals to improve the overall signal quality. The overall signal quality of the multiple signals can refer to the average signal quality of the multiple signals.

[0110] In some embodiments, the inherent polarization transmission effect of MIMO systems can be utilized for signal compensation. Specifically, in the time domain, the multiple signals are transmitted in parallel, and due to channel interference, there is inter-signal crosstalk between them. This inter-signal crosstalk can also be called inter-channel interference. Compensation can be made for the multiple signals exhibiting inter-signal crosstalk. This compensation can be achieved by increasing the non-Gaussianty of the multiple signals. In other words, the greater the non-Gaussianty of the multiple signals, the better their signal quality. The non-Gaussianty of the multiple signals refers to the overall non-Gaussianty of the multiple signals. In one example, the overall non-Gaussianty of the multiple signals is the average of the non-Gaussianty of each individual signal in the multiple signals. In another example, the overall non-Gaussianty of the multiple signals is the sum of the non-Gaussianty of each individual signal in the multiple signals.

[0111] In one example of this embodiment, the compensation model used in step 302 is based on the ICA algorithm.

[0112] In one illustrative example, the compensation model used in step 302 consists of multiple tap matrices. A tap matrix can also be called a filter, and signal compensation can also be called signal filtering. The number of tap matrices constituting the compensation model is equal to the product of the number of transmitting ports at the transmitting end and the number of receiving ports at the receiving end.

[0113] See Figure 4 Taking the X signal and Y signal (each of the X signal and Y signal is a signal) as an example for signal compensation, the compensation model is shown in Equation (1).

[0114]

[0115] Where M′ represents a matrix composed of multiple tap matrices (i.e., the compensation model), H represents the parameter in the tap matrix used to represent crosstalk between signals, L is the difference obtained by subtracting 1 from the length of the input data required by the tap matrix, L is an integer greater than 1, and L is an even number; M represents the tap matrix; x represents the X signal in the multiple signals, and y represents the Y signal in the multiple signals. H is also called the tap coefficient.

[0116] Additionally, the subscript ab indicates that the input is a and the output is b, where a and b each represent one of x and y.

[0117] A tap matrix, also known as a filter, can be used to equalize and demultiplex the input signal. Multiple tap matrices forming the compensation model constitute a butterfly filter result. Each filter's tap coefficients are pairwise correlated with transceiver impairments, thus aiding in transceiver impairment separation. Here, "transceiver" refers to the receiver and transmitter, and "transceiver impairments" refers to both receiver and transmitter impairments.

[0118] When the compensation model consists of multiple tap matrices, in step 302, a vector matrix of multiple signals can be constructed. Then, using these multiple tap matrices, the vector matrix of the multiple signals is calculated to obtain the signal compensation result of the multiple signals. Taking the tap matrix shown in equation (1) as an example, the signal compensation result can be characterized by equation (2).

[0119]

[0120] Among them, S OUT The symbol represents the signal compensation result, x represents the Y signal in the multiple signals, y represents the Y signal in the multiple signals, M represents the tap matrix, the superscript * represents the conjugate operation, and the superscript T represents the transpose operation. The vector matrix represents the signals in the multiple signals, and n represents the time index of the signal in the multiple signals.

[0121] In some embodiments, such as Figure 4 As shown, a vector matrix of multiple signals can be constructed using a preprocessing scheme. The preprocessing scheme includes the following components.

[0122] The whitening matrix is ​​obtained by performing eigenvalue decomposition on the multi-channel signal. Eigenvalue decomposition yields an eigenvalue diagonal matrix and an eigenvector matrix. These matrices are used to construct the whitening matrix. Taking the multi-channel signal as X and Y signals as an example, the resulting whitening matrix is ​​shown in equation (3).

[0123]

[0124] Where V represents the whitening matrix, Λ represents the eigenvalue diagonal matrix obtained by eigenvalue decomposition of the multi-channel signals, U represents the eigenvector matrix obtained by eigenvalue decomposition of the multi-channel signals, the superscript T represents the transpose operation, and W represents the parameter in the whitening matrix used to represent crosstalk between signals.

[0125] like Figure 4 As shown, the whitening matrix can be used to whiten the multi-channel signal to obtain the whitened multi-channel signal.

[0126] Taking the X signal and Y signal as an example, the whitening process and the whitened multi-channel signal are shown in Equation (4).

[0127]

[0128] Among them, S IN This represents the signal from the multi-channel signal before whitening. represents the signal in the whitened multi-channel signal, n represents the time index of the signal in the multi-channel signal, x represents the Y signal in the multi-channel signal, and y represents the Y signal in the multi-channel signal.

[0129] Whitening processing can make the mean of different signals zero and the variance one, thus achieving decorrelation between them. Here, correlation refers to a linear relationship. Decorrelation between different signals means eliminating the linear relationship between them.

[0130] After whitening the multiple signals, a vector matrix of the multiple signals can be constructed based on the whitened multiple signals using a time sliding window of L+1 units. Taking the multiple signals as X and Y signals as an example, the vector matrix of the multiple signals can be represented by equation (5).

[0131]

[0132] in, This represents the vector matrix of signals in the whitened multi-channel signal. Other symbols are as described above.

[0133] The feature vector matrix constructed through the preprocessing scheme can facilitate subsequent parameter updates, accelerate the convergence speed of the algorithm, and ensure that the dimension of the input signal matches the subsequent tap matrix.

[0134] In step 302, multiple signals can be compensated to obtain a compensation result. In step 303, the parameters of the compensation model are updated in a way that improves the compensation result.

[0135] See Figure 4 Step 303, also known as parameter estimation or parameter update, aims to estimate or update the parameters that can improve the compensation results to meet preset conditions.

[0136] In some embodiments, the parameters of the compensation model can be randomly updated, and then it can be determined whether the compensation result of the compensation model with updated parameters for multiple signals meets the preset condition. If it does, the parameter updating stops. If it does not, the parameters of the compensation model are randomly updated again, and the compensation result of the compensation model with updated parameters for multiple signals meets the preset condition, and so on, until parameters that can improve the compensation result to meet the preset condition are obtained.

[0137] In some embodiments, as described above, crosstalk between multiple signals is compensated by increasing the non-Gaussianity of the multiple signals. Therefore, in step 303, the parameters of the compensation model are updated with the aim of maximizing the non-Gaussianity of the multiple signals. The parameters of the compensation model are the tap coefficients of the tap matrix that makes up the compensation model. In step 303, the tap coefficients of the filter are updated by maximizing the non-Gaussianity of the multiple signals, and the output signal power is normalized.

[0138] In an illustrative example, taking the multiplexed signals as X and Y, the parameters of the compensation model can be updated as described above.

[0139] The signal compensation result can be normalized by using equation (6) to obtain the normalized signal compensation result. Normalizing the output signal of the compensation model (i.e., the signal compensation result) eliminates the inherent amplitude ambiguity of the ICA algorithm.

[0140]

[0141] in, This represents the normalized signal compensation result, and E represents the average operation.

[0142] In one example, such as Figure 4 As shown, the non-Gaussian property can be maximized using Newton's method. Specifically, as shown in equation (7), equation (7) includes equations (71), (72), (73), and (75).

[0143]

[0144]

[0145]

[0146]

[0147] Where i∈[-L / 2 L / 2], The superscript "+" represents the parameters of the updated compensation model, and the superscript "-" represents the parameters of the original compensation model. The nonlinear function must satisfy three characteristics: second-order differentiability, smoothness, and consistency with the Gaussian properties of the transmitted signal. In one example, the nonlinear function used could be the Sigmoid function.

[0148] The normalized signal (i.e., the signal compensation result) is subjected to nonlinear operation by equation (7) to simulate the probability density distribution of the signal; then, the first and second derivatives of the nonlinear signal are calculated; and then, the non-Gaussianity is maximized by Newton's method to update the tap coefficients of the tap matrix (i.e., the parameters of the compensation model).

[0149] See Figure 4 The updated tap coefficients can be normalized using the symmetric decorrelation method through equation (8), while ensuring the orthogonality of the X and Y signals.

[0150] M′=(M′M′ H ) -0.5 M′ (8)

[0151] The superscript H represents the conjugate transpose operation.

[0152] in addition, Figure 4 In the middle, H x Represents H xx and H xy H y Represents H yx and H yy .

[0153] In another example, the natural gradient iteration method can be used to maximize the non-Gaussian property.

[0154] Thus, in step 303, the parameters of the compensation model can be updated.

[0155] It can be understood that step 303 is executed in multiple rounds. After each execution of step 303, it can be determined whether the compensation result of the compensation model with updated parameters for multiple signals meets preset conditions. If the preset conditions are not met, step 303 is executed again. If the preset conditions are met, step 304 is executed, that is, when the compensation result meets the preset conditions, the parameters of the compensation model are obtained. The parameters of the compensation model obtained in step 304 refer to the current parameters or the parameters after the most recent parameter update.

[0156] In some embodiments, the preset condition is that the compensation result is greater than or equal to a preset threshold. This threshold can be a signal quality threshold. This threshold can be set by the developer based on experience or experimentation.

[0157] In some embodiments, the preset condition is that the compensation result reaches the optimal result achievable by using the compensation model to compensate multiple signals. That is, the optimal result achievable by compensating multiple signals when the parameters of the compensation model can be varied, while other aspects (e.g., structure and algorithm) remain unchanged. Compensation results obtained by using a compensation model with parameters other than those corresponding to the optimal result for the multiple signals are all less than or equal to the optimal result.

[0158] Next, in step 305, the parameters of the compensation model obtained in step 304 can be used to identify damage to the network system. Specifically, the parameters of the compensation model obtained in step 304 can be used to identify physical layer damage to the network system, that is, whether the physical layer devices (i.e., hardware) of the network system have been damaged.

[0159] In some embodiments, the parameters of the compensation model obtained in step 304 can be used to determine whether the network system has been damaged. When it is determined that the network system has been damaged, the damage to the network system is confirmed.

[0160] In some embodiments, the location of the impairment in the network system can be identified based on the parameters of the compensation model obtained in step 304. This location includes the receiver and / or the transmitter. Specifically, it identifies whether the impairment occurs at the receiver, the transmitter, or both simultaneously.

[0161] In some embodiments, the degree of damage, i.e. the severity of damage, can be identified based on the parameters of the compensation model obtained in step 304.

[0162] In some embodiments, the type of damage can be identified based on the parameters of the compensation model obtained in step 304; wherein the type of damage includes polarization imbalance damage. That is, based on the parameters of the compensation model obtained in step 304, it can be identified whether the damage is polarization imbalance damage.

[0163] Polarization imbalance impairment, also known as channel imbalance impairment, includes clock skew and amplitude imbalance. Clock skew refers to timing errors between different channels and is a typical physical layer device impairment. Amplitude imbalance refers to power differences between different channels and is also a typical physical layer device impairment.

[0164] Based on the parameters of the compensation model obtained in step 304, it is possible to identify whether the damage to the network system is clock skew, amplitude imbalance, or a combination of both.

[0165] In some embodiments, step 305 can be implemented as follows.

[0166] In step 305, taking the example of multiple signals consisting of X and Y signals, the process matrix can be obtained by equation (9) based on the parameters in the updated tap matrix, i.e. the parameters of the compensation model obtained in step 304.

[0167]

[0168] Wherein, P represents the process matrix corresponding to the tap matrix, P′ represents the matrix composed of the process matrices corresponding to the multiple tap matrices, and T represents the parameters in the process matrix.

[0169] By combining the process matrix P, the clock offset defect, as well as the location and extent of the clock offset, can be identified through equations (10), (11), and (12). Equation (10) is composed of equations (10-1) and (10-2).

[0170]

[0171]

[0172] τ Tx =(τ xx -τ yx +τ xy ) / 2 (11)

[0173] τ Rx =(τ xx -τ xy +τ yx ) / 2 (12)

[0174] Where τ represents clock skew, g represents amplitude imbalance, Tx represents the transmitting end, Rx represents the receiving end, f represents frequency domain index, and d represents partial derivative. s The value represents the sampling rate, the superscript f indicates that the current variable belongs to the frequency domain, and FFT stands for Fourier transform.

[0175] The absolute value of τ represents the degree of clock skew impairment, and the absolute value of τ is positively correlated with the degree of clock skew impairment. Specifically, τ Tx A non-zero value indicates a clock skew impairment at the transmitting end, and τ Tx The magnitude of the absolute value of τ is positively correlated with the degree of damage caused by the clock skew at the transmitting end. If τ Tx If τ is zero, then no clock skew occurs at the transmitting end. Similarly, τ Rx A non-zero value indicates a clock skew impairment at the receiver, and τ Rx The magnitude of the absolute value of τ is positively correlated with the degree of damage caused by the clock skew at the receiving end. If τ RxIf the value is zero, then the receiving end does not experience clock skew damage.

[0176] By combining the process matrix P, the amplitude imbalance can be identified through equations (13) and (14), as well as the location and extent of the amplitude imbalance.

[0177] g Tx =(E(|P xx |)E(|P xy |) / (E(|P yx |)E(|P yy |))) (13)

[0178] g Rx =(E(|P xx |)E(|P yx |) / (E(|P xy |)E(|P yy |))) (14)

[0179] Here, g represents amplitude imbalance, and other symbols are explained above.

[0180] The absolute value of g represents the degree of impairment caused by amplitude imbalance, and the absolute value of g is positively correlated with the degree of impairment caused by amplitude imbalance. Specifically, g Tx A non-zero value indicates an amplitude imbalance at the transmitting end, and g Tx The magnitude of the absolute value is positively correlated with the degree of impairment caused by amplitude imbalance at the transmitting end. If g Tx If g is zero, then no amplitude imbalance has occurred at the transmitting end. Similarly, g Rx A non-zero value indicates an amplitude imbalance at the receiving end, and g Rx The magnitude of the absolute value is positively correlated with the degree of impairment caused by amplitude imbalance at the receiving end. If g Rx If the value is zero, then no amplitude imbalance has occurred at the receiving end.

[0181] Thus, the above method can be used to identify damage to the network system.

[0182] In some embodiments, step 305, identifying the impairment of the network system based on the acquired parameters of the compensation model, includes: identifying the location of the impairment in the network system based on the acquired parameters of the compensation model, wherein the location includes the receiving end and / or the transmitting end. That is, in step 305, it is identified whether the impairment occurs at the receiving end, at the transmitting end, or whether both the transmitting and receiving ends simultaneously transmitted the impairment.

[0183] In this embodiment, after identifying the location of the damage, step 306 can be executed, sending the acquired parameters of the compensation model to the location of the damage. These parameters are used to pre-compensate the signal at that location. Specifically, the parameters of the compensation model acquired in step 304 are sent to the location identified in step 305, and these acquired parameters are used to pre-compensate the signal at that location. The location of the damage is the location where the damage occurs. As described above, the parameters of the compensation model acquired in step 304 are parameters of a compensation model whose compensation results for multiple signals meet preset conditions. Using these parameters to perform pre-compensation at the location of the damage can improve the speed of signal compensation and ensure that the signal is compensated to meet preset conditions in advance, overcoming the damage caused by network impairments. Pre-compensation includes, but is not limited to, adjusting hardware parameters.

[0184] In an illustrative example, such as Figure 2 As shown, the receiving end includes an Rx damage pre-compensation module. If damage is detected at the receiving end, the parameters of the compensation model obtained in step 304 can be sent to the Rx damage pre-compensation module. The Rx damage pre-compensation module can pre-compensate the signal received by the receiving end based on the parameters of the compensation model obtained in step 304.

[0185] In an illustrative example, such as Figure 2 As shown, the receiving end includes a Tx damage pre-compensation module. If damage is detected at the transmitting end, the parameters of the compensation model obtained in step 304 can be sent to the Tx damage pre-compensation module. The Tx damage pre-compensation module can pre-compensate the signal to be transmitted by the transmitting end based on the parameters of the compensation model obtained in step 304.

[0186] In an illustrative example, parameter transfer can be achieved through a return channel. For instance, additional wavelengths, frequency bands, time slots, data code blocks, and data frames can be introduced into the return channel as carriers of impairment information to achieve parameter return. As mentioned above, Figure 3 The method shown can be executed by the analysis module. The analysis module obtains the parameters of the compensation model and the identified damage location through step 304, and sends the parameters to the damage location via the feedback channel. These parameters can be used for pre-compensation of the signal at the damage location.

[0187] The network impairment identification method provided in this application can be widely applied in various fields, such as coherent optical fiber communication, mode-division multiplexing optical fiber communication, wireless communication, speech signal processing, and biomedical signal processing, which employ MIMO architectures, to achieve blind source separation and delay estimation between different signal sources. Adjustments can be made as follows when applying this method in different fields.

[0188] 1. The dimensions of the compensation model algorithm architecture can be changed according to the number of information sources (i.e., transmitting ports) and the number of receivers (i.e., receiving ports). Let the number of information sources and the number of receivers be n and m respectively (m≥n must be satisfied). Then, the preprocessing branch changes from the current 2 branches to m, the parameter estimation branch changes from the current 2 branches to n, the dimension of the tap matrix M′ changes from the current 2×2 block matrix (each block is of length L) to an m×n block matrix, and the resulting process matrix P′ also changes from a 2×2 block matrix to an m×n block matrix. Finally, the same parameter extraction method can be used to extract the amplitude imbalance and clock offset between different branches of the transmitting and receiving ends from the process matrix P′.

[0189] 2. Select a nonlinear function that matches the Gaussian characteristics of the transmitted signal. The Gaussian characteristics of signals transmitted in different technical fields vary, necessitating the selection of a nonlinear function that matches the Gaussian characteristics of the transmitted signal.

[0190] In summary, the network impairment identification method provided in this application can identify network impairments without requiring additional hardware in the network system or knowledge of the signal format. This method is low-cost, has no impact on communication performance, is format-transparent, and has a wide range of applications.

[0191] More specifically, the network impairment identification method provided in this application constructs a butterfly filter algorithm architecture that facilitates ICA (Internal Classification Analysis). This is beneficial for utilizing the polarization crosstalk naturally present in network systems to achieve impairment decomposition and monitoring, while simultaneously resolving the amplitude ambiguity problem in ICA. Amplitude ambiguity refers to the fact that the amplitudes of the signals in each branch of the output filter are not the same, exhibiting an uncertain proportional relationship. Resolving the amplitude ambiguity means achieving a uniform amplitude ratio for each output signal, establishing a definite proportional relationship.

[0192] Furthermore, the network impairment identification method provided in this application replaces the signal decision in the traditional scheme with maximizing non-Gaussianity to achieve parameter updates, thus solving the problem that monitoring performance depends on the signal format.

[0193] The network impairment identification method provided in this application integrates all process parameters of the algorithm and separates and monitors transceiver XY clock offset and amplitude imbalance.

[0194] See Figure 5 This application provides a network impairment identification device 500. The device 500 is configured at a receiving end in a network system, which also includes a transmitting end. The receiving end has multiple receiving ports, and the transmitting end has multiple transmitting ports. The number of receiving ports is greater than or equal to the number of transmitting ports in the transmitting end. Figure 5As shown, the device 500 includes:

[0195] The receiving unit 510 is used to receive multiple signals transmitted by the transmitting end through the multiple transmitting ports, wherein the multiple signals correspond one-to-one with the multiple transmitting ports, and one of the multiple signals is received by at least one of the multiple receiving ports.

[0196] The compensation unit 520 is used to compensate the multiple signals using a compensation model to obtain the compensation result of the multiple signals;

[0197] The update unit 530 is used to update the parameters of the compensation model in a direction that improves the compensation result;

[0198] The acquisition unit 540 is used to acquire the parameters of the compensation model when the compensation result meets the preset conditions;

[0199] The identification unit 550 is used to identify damage to the network system based on the parameters of the acquired compensation model.

[0200] The functions of each unit in the network damage identification device can be referred to the above description. Figure 3 The implementation of the method embodiments shown will not be described in detail here.

[0201] This application provides a communication device. See also... Figure 6 The communication device may include a memory 610 and a processor 620; wherein the memory 610 is used to store computer instructions; and the processor 620 is used to execute the computer instructions stored in the memory, enabling the communication device to perform... Figure 3 The method shown.

[0202] It is understood that the processor 620 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. A general-purpose processor can be a microprocessor or any conventional processor.

[0203] This application also provides a chip for performing... Figure 3 The method shown.

[0204] The method steps in the embodiments of this application can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory (RAM), flash memory, read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), registers, hard disks, portable hard disks, CD-ROMs, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can reside in an ASIC.

[0205] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0206] It is understood that the various numerical designations used in the embodiments of this application are merely for descriptive convenience and are not intended to limit the scope of the embodiments of this application.

Claims

1. A method for identifying network impairments, characterized in that, A receiving end is applied in a network system, the network system further comprising a transmitting end, wherein the receiving end has multiple receiving ports, the transmitting end has multiple transmitting ports, and the number of receiving ports in the receiving end is greater than or equal to the number of transmitting ports in the transmitting end; the method includes: The multiple signals transmitted by the transmitting end through the multiple receiving ports are received through the multiple receiving ports, wherein the multiple signals correspond one-to-one with the multiple transmitting ports, and one of the multiple signals is received by at least one of the multiple receiving ports. The compensation model is used to compensate the multiple signals to obtain the compensation results of the multiple signals; Update the parameters of the compensation model in a way that improves the compensation result; When the compensation result meets the preset conditions, the parameters of the compensation model are obtained, wherein the parameters of the compensation model reflect the mutual influence between the multiple signals caused by the damage to the network system. Based on the parameters of the acquired compensation model, the damage to the network system is identified.

2. The method according to claim 1, characterized in that, The conditions include: the compensation result is greater than or equal to a preset threshold, or the compensation result reaches the optimal result that can be achieved by using the compensation model to compensate the multi-channel signals.

3. The method according to claim 1 or 2, characterized in that, The step of identifying damage to the network system based on the acquired parameters of the compensation model includes: confirming that damage has occurred in the network system based on the acquired parameters of the compensation model.

4. The method according to claim 1 or 2, characterized in that, The identification of network system impairments based on the acquired parameters of the compensation model includes: Based on the parameters of the acquired compensation model, the location of the damage in the network system is identified, and the location includes the receiving end and / or the sending end; And / or, Based on the parameters of the obtained compensation model, the degree of damage is identified; And / or, Based on the parameters of the acquired compensation model, the type of damage is identified; wherein, the type of damage includes polarization imbalance damage.

5. The method according to claim 1, characterized in that, The multiple signals experience crosstalk during transmission; The compensation for the multi-channel signals includes: increasing the non-Gaussian properties of the multi-channel signals.

6. The method according to claim 5, characterized in that, The step of updating the parameters of the compensation model in order to improve the compensation result includes updating the parameters of the compensation model with the aim of maximizing the non-Gaussianity of the multi-channel signals.

7. The method according to claim 5 or 6, characterized in that, The compensation model includes multiple tap matrices; The process of using a compensation model to compensate the multiple signals and obtaining the compensation results for the multiple signals includes: Construct the vector matrix of the multi-channel signals; Using the multiple tap matrices, the vector matrix of the multiple signals is calculated to obtain the signal compensation result of the multiple signals; The signal compensation result is characterized as follows: in, Represents the signal compensation result. This represents the first signal in the multiple signals. This represents the second signal in the multiple signals. This represents the tap matrix, where the superscript * indicates the conjugate operation and the superscript T indicates the transpose operation. The vector matrix represents the signals in the multiple signals, and n represents the time index of the signals in the multiple signals; The tap matrix is ​​characterized as follows: in, A matrix representing the plurality of tap matrices. The parameter in the tap matrix represents the crosstalk between signals. L is the difference between the length of the input data required by the tap matrix and 1. L is an integer greater than 1 and L is an even number.

8. The method according to claim 7, characterized in that, The vector matrix for constructing the multi-channel signals includes: The multi-channel signals are subjected to eigenvalue decomposition to obtain a whitening matrix; The whitening matrix is ​​used to whiten the multiple signals to obtain the whitened multiple signals. Based on the whitened multi-channel signals, a vector matrix of the multi-channel signals is constructed using a time sliding window of L+1 unit durations. The whitening matrix is ​​represented as follows: in, Represents the whitening matrix. This represents the diagonal matrix of eigenvalues ​​obtained by eigenvalue decomposition of the multiple signals. This represents the eigenvector matrix obtained by eigenvalue decomposition of the multiple signals, with the superscript T indicating the transpose operation. This represents the parameters in the whitening matrix used to represent crosstalk between signals; in, The signal representing the multiple signals before whitening. This represents the signal in the multiplexed signal after whitening; The whitened multi-channel signals are characterized as follows: The vector matrix of the multi-channel signal is represented as follows: in, The vector matrix representing the signals in the whitened multi-channel signal.

9. The method according to claim 8, characterized in that, Updating the parameters of the compensation model in a way that improves the compensation result includes: The signal compensation result is normalized by the first formula to obtain the normalized signal compensation result. The second equation maximizes the non-Gaussian nature of the normalized signal compensation result to update the parameters in the tap matrix; wherein... The first formula includes: ； in, This represents the normalized signal compensation result. Represents average operation; The second formula includes: in, , This represents a nonlinear function. The superscript + indicates the updated parameters, and the superscript - indicates the parameters before the update.

10. The method according to claim 9, characterized in that, Identifying the location of the network impairment in the network system based on the parameters of the acquired compensation model includes: Based on the parameters in the updated tap matrix, the process matrix is ​​obtained; Based on the process matrix, the network impairment is identified using the third equation; The process matrix is ​​characterized as follows: The third formula includes: in, This represents the process matrix corresponding to the tap matrix. The matrix representing the process matrices corresponding to the multiple tap matrices. Represents the parameters in the process matrix. Represents clock offset. This indicates an imbalance in amplitude. Representing the sending end, Representing the receiving end, Represents the frequency domain index. Represents finding the partial derivative. Represents the sampling rate, superscript This indicates that the current variable belongs to the frequency domain. This represents the Fourier transform.

11. The method according to claim 1, characterized in that, The step of identifying the damage to the network system based on the parameters of the acquired compensation model includes: identifying the location of the damage in the network system based on the parameters of the acquired compensation model, wherein the location includes the receiving end and / or the transmitting end; The method further includes: sending the acquired parameters of the compensation model to the location, wherein the acquired parameters of the compensation model are used to pre-compensate the signal at the location.

12. A network damage identification device, characterized in that, A receiving end configured in a network system, the network system further comprising a transmitting end, wherein the receiving end has multiple receiving ports, the transmitting end has multiple transmitting ports, and the number of receiving ports in the receiving end is greater than or equal to the number of transmitting ports in the transmitting end; the device includes: A receiving unit is configured to receive multiple signals transmitted by the transmitting end through the multiple transmitting ports, wherein the multiple signals correspond one-to-one with the multiple transmitting ports, and one of the multiple signals is received by at least one of the multiple receiving ports. The compensation unit is used to compensate the multiple signals using a compensation model to obtain the compensation result of the multiple signals; An update unit is used to update the parameters of the compensation model in a direction that improves the compensation result; The acquisition unit is used to acquire the parameters of the compensation model when the compensation result meets the preset conditions, wherein the parameters of the compensation model reflect the mutual influence between the multiple signals caused by the damage to the network system. An identification unit is used to identify damage to the network system based on the parameters of the acquired compensation model.

13. The apparatus according to claim 12, characterized in that, The conditions include: the compensation result is greater than or equal to a preset threshold, or the compensation result reaches the optimal result that can be achieved by using the compensation model to compensate the multi-channel signals.

14. The apparatus according to claim 12 or 13, characterized in that, The identification unit is used to: confirm that the network system has been damaged based on the parameters of the acquired compensation model.

15. The apparatus according to claim 12 or 13, characterized in that, The identification unit is used for: Based on the parameters of the acquired compensation model, the location of the damage in the network system is identified, and the location includes the receiving end and / or the sending end; And / or, Based on the parameters of the obtained compensation model, the degree of damage is identified; And / or, Based on the parameters of the acquired compensation model, the type of damage is identified; wherein, the type of damage includes polarization imbalance damage.

16. The apparatus according to claim 12, characterized in that, The multiple signals experience inter-signal crosstalk during transmission; the compensation unit is used to increase the non-Gaussian properties of the multiple signals.

17. The apparatus according to claim 16, characterized in that, The updating unit is used to update the parameters of the compensation model with the aim of maximizing the non-Gaussian properties of the multiple signals.

18. A communication device, characterized in that, include: Memory and processor; The memory is used to store computer instructions; the processor is used to execute the computer instructions stored in the memory to implement the method of any one of claims 1-11.

19. A chip, characterized in that, The chip is used to perform the method according to any one of claims 1-11.

20. A computer storage medium, characterized in that, The computer storage medium includes computer instructions that, when executed on a communication device, cause the communication device to perform the method of any one of claims 1-11.

Images