Channel equalizer construction method and device, computer equipment, readable storage medium and program product

Abstract

The invention relates to a channel equalizer construction method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the steps of obtaining a symbol sequence sample; performing sliding window processing on the plurality of symbols according to a symbol receiving sequence by using a sliding window, and for a plurality of current target symbols in the sliding window, inputting the plurality of target symbols into a feed-forward equalizer to obtain feed-forward output; for each feed-forward output after the Lth feed-forward output, acquiring L feed-forward outputs before the current feed-forward output, and inputting the L feed-forward outputs into a feedback equalizer to obtain feedback output; determining a prediction symbol based on the current feed-forward output and the feedback output; and determining a prediction error based on the prediction symbol and the expected response, and adjusting parameters in the feed-forward equalizer and the feedback equalizer based on the prediction error until the prediction error obtained by calculation is smaller than a preset value. The inter-symbol interference of balanced symbols can be reduced.

Description

Technical Field

[0001] The present application relates to the field of communication technology, and in particular to a channel equalizer construction method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Art

[0002] With the rapid development of society, the importance of digital communication technology in the information age is increasing. People's desire for digital communication technology and communication quality is also increasing. Channel equalization technology, as a technology that can effectively improve the transmission performance of communication systems and reduce the interference caused by various non-ideal factors to communication systems, plays an increasingly important role in digital communication systems. Nowadays, channel equalization technology is widely used in various digital communication systems.

[0003] In the related art, a feedforward equalizer is constructed and used to reduce interference. However, the feedforward equalizer cannot eliminate the inter-symbol interference of the equalized symbols. Summary of the invention

[0004] Based on this, it is necessary to provide a channel equalizer construction method, apparatus, computer equipment, computer-readable storage medium and computer program product that can reduce the inter-symbol interference of equalized symbols in order to address the above technical problems.

[0005] In a first aspect, the present application provides a channel equalizer construction method, comprising:

[0006] Acquire a symbol sequence sample, wherein the symbol sequence sample includes a plurality of symbols arranged in sequence according to a symbol reception order;

[0007] Using a sliding window to perform sliding window processing on the multiple symbols according to the symbol receiving order, for the multiple target symbols currently in the sliding window, inputting the multiple target symbols into a feedforward equalizer to obtain a feedforward output;

[0008] For each feedforward output after the Lth feedforward output, obtain the L feedforward outputs before the current feedforward output, and input the L feedforward outputs into the feedback equalizer to obtain a feedback output;

[0009] Determining a prediction sign based on the current feedforward output and the feedback output;

[0010] Based on the predicted symbol and the expected response, a prediction error is determined, and based on the prediction error, parameters in the feedforward equalizer and the feedback equalizer are adjusted until the calculated prediction error is less than a preset value.

[0011] In one embodiment, the size of the sliding window is 2K1, the number of symbols in the symbol sequence sample is N; and the using the sliding window to perform sliding window processing on the multiple symbols according to the symbol reception order includes:

[0012] Extracting the 1st to 2K1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window;

[0013] After obtaining the corresponding feedforward output, extracting the 2nd to 2K1+1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window;

[0014] After the corresponding feedforward output is obtained, the extraction operation is continued until the current last target symbol in the sliding window is the Nth symbol in the symbol sequence sample.

[0015] In one embodiment, the feedforward equalizer includes an input layer, a hidden layer and an output layer; the number of neurons in the input layer is the same as the number of multiple target symbols, and the number of neurons in the hidden layer is the same as the order of the channel that the equalizer needs to fit; the inputting the multiple target symbols into the feedforward equalizer to obtain the feedforward output includes:

[0016] Inputting the plurality of target symbols into the plurality of first neurons in the input layer respectively;

[0017] Each first neuron in the i-th set inputs the received target symbols into the i-th order neuron in the hidden layer; the i-th set is a set consisting of the first neurons corresponding to a preset number of target symbols before the central symbol and a preset number of target symbols after the central symbol among the multiple target symbols;

[0018] Each order of neurons in the hidden layer performs corresponding operations on the received target symbols, obtains the operation results and outputs them to the output layer;

[0019] The output layer adds the received operation results to obtain a feedforward output.

[0020] In one embodiment, each order of neurons in the hidden layer performs corresponding operations on the received target symbols, including:

[0021] For the i-th order neuron, obtain the i-th tap coefficient corresponding to each received target symbol;

[0022] The i-th order tap coefficient corresponding to each target symbol is multiplied by the corresponding target symbol, the multiplication results are summed, the summed result and the bias corresponding to the i-th order neuron are added, the i-th power of the summed result is obtained, and the operation result is obtained.

[0023] In one embodiment, inputting the L times feedforward output into a feedback equalizer to obtain a feedback output comprises:

[0024] Get the decision coefficient corresponding to the i-th feedforward output before the current feedforward output;

[0025] Multiply the decision coefficient corresponding to the i-th feedforward output by the i-th feedforward output to obtain the decision result corresponding to the i-th feedforward output;

[0026] The decision results corresponding to the L feedforward outputs are added together to obtain the feedback output.

[0027] In one embodiment, the parameters of the feedforward equalizer include tap coefficients, and the parameters of the feedback equalizer include decision coefficients; and adjusting the parameters in the feedforward equalizer and the feedback equalizer based on the prediction error includes:

[0028] Obtaining the mean square error gradient of the prediction error;

[0029] For the tap coefficient or the decision coefficient, obtain the current value of the current coefficient, use the learning rate to multiply the mean square error gradient to obtain the multiplication result; use the current value to subtract the multiplication result, and use the subtraction result to update the current value.

[0030] In a second aspect, the present application also provides a channel equalizer construction device, including:

[0031] An acquisition module, used for acquiring a symbol sequence sample, wherein the symbol sequence sample comprises a plurality of symbols arranged in sequence according to a symbol reception order;

[0032] A sliding window processing module, configured to perform sliding window processing on the multiple symbols according to the symbol receiving order using a sliding window, and for multiple target symbols currently in the sliding window, input the multiple target symbols into a feedforward equalizer to obtain a feedforward output;

[0033] The acquisition module is further used to acquire L feedforward outputs before the current feedforward output for each feedforward output after the Lth feedforward output, input the L feedforward outputs to the feedback equalizer to obtain feedback output; and determine the prediction symbol based on the current feedforward output and the feedback output;

[0034] A parameter adjustment module is used to determine a prediction error based on the predicted symbol and the expected response, and adjust parameters in the feedforward equalizer and the feedback equalizer based on the prediction error until the calculated prediction error is less than a preset value.

[0035] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0036] Acquire a symbol sequence sample, wherein the symbol sequence sample includes a plurality of symbols arranged in sequence according to a symbol reception order;

[0037] Using a sliding window to perform sliding window processing on the multiple symbols according to the symbol receiving order, for the multiple target symbols currently in the sliding window, inputting the multiple target symbols into a feedforward equalizer to obtain a feedforward output;

[0038] For each feedforward output after the Lth feedforward output, obtain the L feedforward outputs before the current feedforward output, and input the L feedforward outputs into the feedback equalizer to obtain a feedback output;

[0039] Determining a prediction sign based on the current feedforward output and the feedback output;

[0040] Based on the predicted symbol and the expected response, a prediction error is determined, and based on the prediction error, parameters in the feedforward equalizer and the feedback equalizer are adjusted until the calculated prediction error is less than a preset value.

[0041] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0042] Acquire a symbol sequence sample, wherein the symbol sequence sample includes a plurality of symbols arranged in sequence according to a symbol reception order;

[0043] Using a sliding window to perform sliding window processing on the multiple symbols according to the symbol receiving order, for the multiple target symbols currently in the sliding window, inputting the multiple target symbols into a feedforward equalizer to obtain a feedforward output;

[0044] For each feedforward output after the Lth feedforward output, obtain the L feedforward outputs before the current feedforward output, and input the L feedforward outputs into the feedback equalizer to obtain a feedback output;

[0045] Determining a prediction sign based on the current feedforward output and the feedback output;

[0046] Based on the predicted symbol and the expected response, a prediction error is determined, and based on the prediction error, parameters in the feedforward equalizer and the feedback equalizer are adjusted until the calculated prediction error is less than a preset value.

[0047] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0048] Acquire a symbol sequence sample, wherein the symbol sequence sample includes a plurality of symbols arranged in sequence according to a symbol reception order;

[0049] Using a sliding window to perform sliding window processing on the multiple symbols according to the symbol receiving order, for the multiple target symbols currently in the sliding window, inputting the multiple target symbols into a feedforward equalizer to obtain a feedforward output;

[0050] For each feedforward output after the Lth feedforward output, obtain the L feedforward outputs before the current feedforward output, and input the L feedforward outputs into the feedback equalizer to obtain a feedback output;

[0051] Determining a prediction sign based on the current feedforward output and the feedback output;

[0052] Based on the predicted symbol and the expected response, a prediction error is determined, and based on the prediction error, parameters in the feedforward equalizer and the feedback equalizer are adjusted until the calculated prediction error is less than a preset value.

[0053] The channel equalizer construction method, device, computer equipment, computer-readable storage medium and computer program product obtain a symbol sequence sample, wherein the symbol sequence sample includes a plurality of symbols arranged in sequence according to the symbol reception order; a sliding window is used to perform sliding window processing on the plurality of symbols according to the symbol reception order; for the current plurality of target symbols in the sliding window, the plurality of target symbols are input into the feedforward equalizer to obtain a feedforward output; for each feedforward output after the Lth feedforward output, the L feedforward outputs before the current feedforward output are obtained, and the L feedforward outputs are input into the feedback equalizer to obtain a feedback output; based on the current feedforward output and the feedback output, a predicted symbol is determined; based on the predicted symbol and the expected response, a prediction error is determined, and the parameters in the feedforward equalizer and the feedback equalizer are adjusted based on the prediction error until the calculated prediction error is less than a preset value. This construction method of introducing a feedback equalizer enables the final channel equalizer to eliminate the inter-symbol interference of the equalized symbols, and restore the received symbols to symbols that are closer to the real symbols sent by the transmitter, with higher accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0055] FIG1 is a diagram of an application environment of a channel equalizer construction method according to an embodiment;

[0056] Figure 2 A schematic diagram of a flow chart of a method for constructing a channel equalizer in one embodiment;

[0057] Figure 3 A schematic diagram of the construction principle of a channel equalizer in one embodiment;

[0058] Figure 4 A schematic diagram of a practical application of a channel equalizer in an embodiment;

[0059] Figure 5 A schematic diagram showing a comparison of bit error rates of various channel equalizers in one embodiment;

[0060] Figure 6 A schematic diagram of a comparison of net rates of various channel equalizers in one embodiment;

[0061] Figure 7 A structural block diagram of a channel equalizer construction device in one embodiment;

[0062] FIG. 8 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0064] The channel equalizer construction method provided in the embodiment of the present application can be applied to the application environment shown in Figure 1. The application environment shown in Figure 1 includes a transmitter and a receiver. After the receiver receives the symbol sent by the transmitter through the channel, the symbol can be input into a pre-constructed channel equalizer, and the channel equalizer restores the symbol to obtain a restored signal. The channel equalizer is constructed in the following manner: obtaining a symbol sequence sample, the symbol sequence sample includes a plurality of symbols arranged in sequence according to the symbol receiving order; using a sliding window to perform sliding window processing on the plurality of symbols according to the symbol receiving order, for the current plurality of target symbols in the sliding window, the plurality of target symbols are input into a feedforward equalizer to obtain a feedforward output; for each feedforward output after the Lth feedforward output, the L feedforward outputs before the current feedforward output are obtained, the L feedforward outputs are input into a feedback equalizer to obtain a feedback output; based on the current feedforward output and the feedback output, a predicted symbol is determined; based on the predicted symbol and the expected response, a prediction error is determined, and based on the prediction error, the parameters in the feedforward equalizer and the feedback equalizer are adjusted until the calculated prediction error is less than a preset value, and the combination of the feedforward equalizer and the feedback equalizer at this time is used as a trained channel equalizer. Since the feedback equalizer is added in the construction process of the channel equalizer, the restored signal can eliminate the inter-symbol interference of the equalized symbol, and is closer to the real symbol sent by the transmitter, and has higher accuracy.

[0065] The transmitting end may be a base station, and the receiving end may be a terminal. This is only one possible scenario, and the transmitting end and the receiving end may also be corresponding devices in other scenarios, which is not limited in the embodiments of the present application.

[0066] The terminal 102 may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, IoT devices, and portable wearable devices. IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, projection devices, etc. Portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. Head-mounted devices may be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc.

[0067] In an exemplary embodiment, Figure 2 As shown, a channel equalizer construction method is provided, and the method is applied to the receiving end in FIG. 1 as an example for description, including the following steps 202 to 210. Among them:

[0068] Step 202: Acquire a symbol sequence sample, where the symbol sequence sample includes a plurality of symbols arranged in sequence according to a symbol reception order.

[0069] Optionally, the symbol sequence sample includes N symbols arranged in sequence according to the symbol reception order.

[0070] Optionally, the multiple symbols in the symbol sequence sample may be received by the receiving end during an actual working process.

[0071] Step 204: Use a sliding window to perform sliding window processing on the multiple symbols according to the symbol reception order, and for the current multiple target symbols in the sliding window, input the multiple target symbols into a feedforward equalizer to obtain a feedforward output.

[0072] Optionally, the size of the sliding window may be set according to actual conditions. Exemplarily, the size of the sliding window may be 2K1, that is, the sliding window may cover 2K1 symbols.

[0073] Optionally, according to the order of symbol reception, the first coincident group may be covered by a sliding window, and multiple symbols in the symbol group may be used as multiple current target symbols in the sliding window, and subsequent processing may be performed; then the sliding window may be moved with a step length of one symbol, that is, the second coincident group may be covered by a sliding window, and multiple symbols in the symbol group may be used as multiple current target symbols in the sliding window, and subsequent processing may be performed; and so on, until the last symbol in the symbol group covered by the sliding window is the last symbol in the symbol sequence sample. The number of symbols in the coincident group is the same as the size of the moving window.

[0074] Optionally, the feedforward equalizer includes an input layer, a hidden layer, and an output layer. For multiple target symbols currently in the sliding window, based on the multiple target symbols, the feedforward output can be obtained by sequentially processing the input layer, the hidden layer, and the output layer. For each position of the sliding window, a corresponding feedforward output can be calculated.

[0075] Step 206: for each feedforward output after the Lth feedforward output, obtain the L feedforward outputs before the current feedforward output, and input the L feedforward outputs to the feedback equalizer to obtain a feedback output.

[0076] Among them, L is a value pre-set according to the actual situation.

[0077] Optionally, a corresponding feedforward output can be calculated at each position of the sliding window. As the sliding window moves, after L feedforward outputs are calculated, for each subsequent feedforward output, the L feedforward outputs before the current feedforward output can be obtained, and the L feedforward outputs are input into the feedback equalizer to obtain the feedback output.

[0078] Step 208: Determine a predicted sign based on the current feedforward output and the feedback output.

[0079] Optionally, the current feedforward output and feedback output may be added to obtain the predicted symbol. Alternatively, the feedforward weight and feedback weight may be preset, the current feedforward output may be multiplied by the feedforward weight, the feedback output may be multiplied by the feedback weight, and the two multiplication results may be added to obtain the predicted symbol.

[0080] Step 210, determining a prediction error based on the predicted symbol and the expected response, and adjusting parameters in the feedforward equalizer and the feedback equalizer based on the prediction error until the calculated prediction error is less than a preset value.

[0081] The expected response is an undistorted symbol of an intermediate symbol of multiple target symbols corresponding to the current feedforward output.

[0082] The prediction error can be obtained by subtracting the predicted symbol from the expected response. The parameters in the feedforward equalizer and the feedback equalizer are adjusted based on the prediction error until the calculated prediction error is less than a preset value. The combination of the feedforward equalizer and the feedback equalizer at this time is used as a trained channel equalizer.

[0083] As described above, a corresponding feedforward output can be calculated at each position of the sliding window. As the sliding window moves, after L feedforward outputs are calculated, for each subsequent feedforward output, the L feedforward outputs before the current feedforward output can be obtained, and the L feedforward outputs are input into the feedback equalizer to obtain the feedback output corresponding to the current feedforward output, and then the prediction symbol, prediction error, adjustment parameters and other actions are calculated. That is to say, after L feedforward outputs are calculated, a parameter adjustment action can be performed for each subsequent feedforward output.

[0084] In the above embodiment, a symbol sequence sample is obtained, and the symbol sequence sample includes a plurality of symbols arranged in sequence according to the symbol reception order; a sliding window is used to perform sliding window processing on the plurality of symbols according to the symbol reception order, and for the current plurality of target symbols in the sliding window, the plurality of target symbols are input into the feedforward equalizer to obtain a feedforward output; for each feedforward output after the Lth feedforward output, the L feedforward outputs before the current feedforward output are obtained, and the L feedforward outputs are input into the feedback equalizer to obtain a feedback output; based on the current feedforward output and the feedback output, a predicted symbol is determined; based on the predicted symbol and the expected response, a prediction error is determined, and based on the prediction error, the parameters in the feedforward equalizer and the feedback equalizer are adjusted until the calculated prediction error is less than a preset value. This construction method of introducing a feedback equalizer enables the final channel equalizer to eliminate the inter-symbol interference of the equalized symbols, and restore the received symbols to symbols that are closer to the real symbols sent by the transmitter, with higher accuracy.

[0085] In some embodiments, the size of the sliding window is 2K1, and the number of symbols in the symbol sequence sample is N; the step of using the sliding window to perform sliding window processing on the multiple symbols in the symbol reception order includes: extracting the 1st to 2K1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; after obtaining the corresponding feedforward output, extracting the 2nd to 2K1+1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; after obtaining the corresponding feedforward output, continuing to perform the extraction operation until the current last target symbol in the sliding window is the Nth symbol in the symbol sequence sample.

[0086] The symbol sequence sample includes N symbols arranged in sequence according to the symbol reception order.

[0087] Optionally, the use process of the sliding window is as follows: extract the 1st to 2K1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window, input these multiple target symbols into the feedforward equalizer to obtain feedforward output; then, extract the 2nd to 2K1+1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window, input these multiple target symbols into the feedforward equalizer to obtain feedforward output; and so on, until the current last target symbol in the sliding window is the Nth symbol in the symbol sequence samples, input the current multiple target symbols in the sliding window into the feedforward equalizer to obtain feedforward output, and the sliding processing ends.

[0088] In the above embodiment, the sliding processing process is introduced. This sliding processing method can support the channel equalizer to complete multiple trainings, thereby improving the training efficiency.

[0089] In some embodiments, the feedforward equalizer includes an input layer, a hidden layer and an output layer; the number of neurons in the input layer is the same as the number of multiple target symbols, and the number of neurons in the hidden layer is the same as the order of the channel that the equalizer needs to fit; the multiple target symbols are input into the feedforward equalizer to obtain a feedforward output, including: inputting the multiple target symbols into multiple first neurons in the input layer respectively; each first neuron in the i-th set inputs the received target symbols into the i-th order neuron in the hidden layer; the i-th set is a set of first neurons corresponding to a preset number of target symbols before the center symbol and a preset number of target symbols after the center symbol among the multiple target symbols; each order of neurons in the hidden layer performs corresponding operations on the received target symbols, obtains the operation results and outputs them to the output layer; the output layer adds the received operation results to obtain the feedforward output.

[0090] For example, see Figure 3 As shown, the order of the channel that the scale needs to fit is 3rd order, and the hidden layer includes three neurons, namely, 1st order neurons, 2nd order neurons and 3rd order neurons. The multiple target symbols can be input into the multiple first neurons in the input layer respectively, and the intermediate symbol of the multiple target symbols is v k , the middle symbol v k The previous K1 symbols are v k-1 、v k-2 、v k-3 ,……v k-K1 , the middle symbol v k The next K1 symbols are v k+1 、v k+2 、v k+3 , ……v k+K1 , the first set can be the intermediate symbol v k The first neurons corresponding to the previous K1 symbols and the next K1 symbols are composed of a set. Each first neuron in the first set inputs the received target symbol into the first-order neuron; the second set can be the intermediate symbol v k The first neurons corresponding to the previous m2 symbols and the next m2 symbols are composed of a set. Each first neuron in the second set inputs the received target symbol into the second-order neuron; the third set can be the intermediate symbol v k The first neurons corresponding to the previous m3 symbols and the next m3 symbols are respectively composed of a set, and each first neuron in the third set inputs the received target symbol into the third-order neuron. The first-order neuron performs a corresponding operation on the received target symbol, obtains the operation result and outputs it to the output layer; similarly, the second-order neuron performs a corresponding operation on the received target symbol, obtains the operation result and outputs it to the output layer; the third-order neuron performs a corresponding operation on the received target symbol, obtains the operation result and outputs it to the output layer; the output layer adds the received operation results to obtain the feedforward output y(k).

[0091] In the above embodiment, the number of neurons in the hidden layer is determined by the order of the channel that the equalizer needs to fit. This structure simplifies the process of selecting structural hyperparameters in the neural network.

[0092] In some embodiments, the step of each order neuron in the hidden layer performing corresponding operations on the received target symbols includes: for the i-th order neuron, obtaining the i-th tap coefficient corresponding to each received target symbol; multiplying the i-th order tap coefficient corresponding to each target symbol with the corresponding target symbol, summing the multiplication results, adding the summed result and the bias corresponding to the i-th order neuron, taking the i-th power of the addition result, and obtaining the operation result.

[0093] Exemplary, combined Figure 3 In the feedforward equalizer shown, the order of the channel that the equalizer needs to fit is 3rd order, and the hidden layer includes three neurons, namely, 1st order neurons, 2nd order neurons and 3rd order neurons. Each first neuron in the 1st set inputs the received target symbol to the 1st order neuron; each first neuron in the 2nd set inputs the received target symbol to the 2nd order neuron; each first neuron in the 3rd set inputs the received target symbol to the 3rd order neuron. The 1st order neuron performs corresponding operations on the received target symbol, obtains the operation result and outputs it to the output layer; similarly, the 2nd order neuron performs corresponding operations on the received target symbol, obtains the operation result and outputs it to the output layer; the 3rd order neuron performs corresponding operations on the received target symbol, obtains the operation result and outputs it to the output layer; the output layer adds the received operation results to obtain the feedforward output y(k).

[0094] Among them, the calculation result of the first-order neuron can be expressed as follows:

[0095] ;

[0096] in, is the intermediate symbol v among multiple target symbols received by the first-order neuron k is the kith target symbol of the benchmark, for The corresponding first-order tap coefficient, b1 is the bias. K1 in the above text.

[0097] The calculation result of the second-order neuron can be expressed as follows:

[0098]

[0099] in, is the intermediate symbol v among multiple target symbols received by the second-order neuron k is the kith target symbol of the benchmark, for The corresponding second-order tap coefficient, b2 is the bias. m2 is less than K1.

[0100] The calculation result of the third-order neuron can be expressed as follows:

[0101]

[0102] in, is the intermediate symbol v among multiple target symbols received by the third-order neuron k is the kith target symbol of the benchmark, for The corresponding third-order tap coefficient, b3 is the bias. m3 is smaller than m2.

[0103] In the above embodiment, the operation process of the i-th order neuron is introduced, and a custom activation function is used to construct the cross-beat frequency term, thereby avoiding redundant high-order nonlinear terms.

[0104] In some embodiments, the step of inputting the L feedforward outputs into a feedback equalizer to obtain a feedback output includes: obtaining a decision coefficient corresponding to the i-th feedforward output before the current feedforward output; multiplying the decision coefficient corresponding to the i-th feedforward output by the i-th feedforward output to obtain a decision result corresponding to the i-th feedforward output; and adding the decision results corresponding to each of the L feedforward outputs to obtain a feedback output.

[0105] Among them, the feedback output can be expressed as follows:

[0106]

[0107] in, is the i-th feedforward output before the current feedforward output, for The corresponding determination coefficient.

[0108] In the above embodiment, a method for obtaining feedback output is introduced. For each feedforward output after the Lth feedforward output, the decision coefficient corresponding to the i-th feedforward output before the current feedforward output can be obtained; the decision coefficient corresponding to the i-th feedforward output is multiplied by the i-th feedforward output to obtain the decision result corresponding to the i-th feedforward output; the decision results corresponding to the L feedforward outputs are added to obtain the feedback output. In the process of constructing a channel equalizer, a feedback equalizer is introduced so that the final channel equalizer can eliminate the inter-symbol interference of the equalized symbols and restore the received symbols to symbols that are closer to the real symbols sent by the transmitter, with higher accuracy.

[0109] In some embodiments, the parameters of the feedforward equalizer include tap coefficients, and the parameters of the feedback equalizer include decision coefficients; adjusting the parameters in the feedforward equalizer and the feedback equalizer based on the prediction error includes: obtaining the mean square error gradient of the prediction error; for the tap coefficient or the decision coefficient, obtaining the current value of the current coefficient, multiplying the learning rate by the mean square error gradient to obtain a multiplication result; subtracting the multiplication result from the current value, and updating the current value using the subtraction result.

[0110] Optionally, the prediction symbol can be expressed as:

[0111]

[0112] in, Represents the prediction symbol.

[0113] Optionally, the prediction error can be expressed as:

[0114]

[0115] in, represents the prediction error, Indicates the expected response.

[0116] Optionally, the mean squared gradient of the prediction error can be expressed as follows:

[0117] Δ J = ∂ e 2 ( k ) ∂ w =− 2 [ d ( k ) − y ( k )] ∂ y ( k ) ∂ w

[0118] for , , , These coefficients can be updated using chain rules. With For example:

[0119] w ff 2 ( k + 1 , i ) = w ff 2 ( k , i ) − μ Δ J = w ff 2 ( k , i ) − μ { − 2 [ d ( k ) − y ( k )] ⋅ 2 [ ∑ i =− m 2 m 2 w ff 2 ( i ) v ( k − i ) + b 2 ] ⋅ v ( k − i ) } in, for The current value of is the product of the learning rate and the mean square error gradient, using renew , it can be achieved The parameters in the feedforward equalizer include the above , , , the parameters in the feedback equalizer include .

[0120] In the above embodiment, a specific method for updating the parameters in the feedforward equalizer and the feedback equalizer is provided. This construction method of introducing the feedback equalizer enables the final channel equalizer to eliminate the inter-symbol interference of the equalized symbols, and restore the received symbols to symbols that are closer to the actual symbols sent by the transmitter, with higher accuracy.

[0121] In some embodiments, the channel equalizer provided in the embodiment of the present application adds a Volterra series, and the channel equalizer provided in the embodiment of the present application is also called a Volterra-like neural network decision feedback equalizer (VINN DFE). Figure 4As shown, at the transmitter, the PAM-8 symbol is generated by an arbitrary waveform generator (AWG) with a 224 GSa / s 3dB bandwidth of about 65GHz. The analog signal is loaded into a thin-film lithium niobate Mach-Zehnder modulator (TFLN MZM) with a 3dB bandwidth of 60GHz and a DC half-wave voltage of 1.3V. The external cavity laser (ECL) operates at a wavelength of 1552.5nm and an optical power of 13dBm. The output optical power of the TFLN-MZM is -5dBm. After 500m standard single-mode fiber (SSMF) transmission, an erbium-doped fiber amplifier (EDFA) is used to compensate for the link loss and the optical signal is boosted to 6dBm by gain control. The received optical power (ROP) is adjusted by a variable optical attenuator (VOA) before entering the photodetector (PD). Finally, the electrical signal output of the PD is extracted by a 59GHz oscilloscope operating at a sampling rate of 256GSa / s. The 10dB bandwidth of the system is 58GHz. Figure 5 The relationship between BER and baud rate in the range of 120 to 130 GBaud is shown. The optical signal ROP is 6 dBm. Since the system frequency response of 10 dB bandwidth is 58 GHz, the BER performance deteriorates rapidly with the increase of baud rate. When the complexity is limited to 100 MACs, the Volterra-like neural network feedforward equalizer (VINN) outperforms the Volterra nonlinear equalizer (VNLE), and the Volterra-like neural network decision feedback equalizer (VINN DFE) outperforms the Volterra nonlinear decision feedback equalizer (VDFE). The Volterra-like neural network feedforward equalizer (VINN) and VDFE have similar performance, showing that the nonlinear decision feedback compensates for the nonlinearity of the feedforward filter. The calculated net rate of attachment Figure 6 As shown. The system net rate is calculated based on the bit rate corresponding to the hard decision BER. As the baud rate increases, the net rate first increases and then decreases. Compared with the VINN DFE provided in the embodiment of the present application, other equalizers are more affected by bandwidth limitations and nonlinearity, and the net rate decreases as the baud rate increases.

[0122] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0123] Based on the same inventive concept, the embodiment of the present application also provides a channel equalizer construction device for implementing the channel equalizer construction method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more channel equalizer construction device embodiments provided below can refer to the limitations of the channel equalizer construction method above, and will not be repeated here.

[0124] In an exemplary embodiment, Figure 7 As shown, a channel equalizer construction device is provided, and the channel equalizer construction device can be a receiver, including:

[0125] An acquisition module 701 is used to acquire a symbol sequence sample, where the symbol sequence sample includes a plurality of symbols arranged in sequence according to a symbol reception order;

[0126] A sliding window processing module 702 is used to perform sliding window processing on the multiple symbols according to the symbol reception order using a sliding window, and for multiple target symbols currently in the sliding window, input the multiple target symbols into a feedforward equalizer to obtain a feedforward output;

[0127] The acquisition module 701 is further configured to acquire L feedforward outputs before the current feedforward output for each feedforward output after the Lth feedforward output, input the L feedforward outputs to the feedback equalizer to obtain a feedback output; and determine a prediction symbol based on the current feedforward output and the feedback output;

[0128] The parameter adjustment module 703 is used to determine the prediction error based on the predicted symbol and the expected response, and adjust the parameters in the feedforward equalizer and the feedback equalizer based on the prediction error until the calculated prediction error is less than a preset value.

[0129] In some embodiments, the size of the sliding window is 2K1, and the number of symbols in the symbol sequence sample is N; the sliding window processing module 702 is used to extract the 1st to 2K1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; after obtaining the corresponding feedforward output, extract the 2nd to 2K1+1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; after obtaining the corresponding feedforward output, continue to perform the extraction operation until the current last target symbol in the sliding window is the Nth symbol in the symbol sequence sample.

[0130] In some embodiments, the feedforward equalizer includes an input layer, a hidden layer and an output layer; the number of neurons in the input layer is the same as the number of multiple target symbols, and the number of neurons in the hidden layer is the same as the order of the channel that the equalizer needs to fit; a sliding window processing module 702 is used to input the multiple target symbols into multiple first neurons in the input layer respectively; each first neuron in the i-th set inputs the received target symbols into the i-th order neuron in the hidden layer; the i-th set is a set of first neurons corresponding to a preset number of target symbols before the center symbol and a preset number of target symbols after the center symbol among the multiple target symbols; each order of neurons in the hidden layer performs corresponding operations on the received target symbols, obtains the operation results and outputs them to the output layer; the output layer adds the received operation results to obtain a feedforward output.

[0131] In some embodiments, the sliding window processing module 702 is used to obtain the i-th tap coefficient corresponding to each received target symbol for the i-th order neuron; multiply the i-th order tap coefficient corresponding to each target symbol with the corresponding target symbol, sum the multiplication results, add the sum result and the bias corresponding to the i-th order neuron, and take the i-th power of the addition result to obtain the calculation result.

[0132] In some embodiments, the acquisition module 701 is used to obtain the decision coefficient corresponding to the i-th feedforward output before the current feedforward output; multiply the decision coefficient corresponding to the i-th feedforward output by the i-th feedforward output to obtain the decision result corresponding to the i-th feedforward output; add the decision results corresponding to the L feedforward outputs to obtain the feedback output.

[0133] In some embodiments, the parameter adjustment module 703 is used to obtain the mean square error gradient of the prediction error; for the tap coefficient or the decision coefficient, obtain the current value of the current coefficient, multiply the learning rate by the mean square error gradient to obtain the multiplication result; subtract the multiplication result from the current value, and use the subtraction result to update the current value.

[0134] Each module in the above-mentioned channel equalizer construction device can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0135] In an exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be shown in FIG8. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as parameters in a feedforward equalizer and a feedback equalizer. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a channel equalizer construction method is implemented.

[0136] Those skilled in the art will understand that the structure shown in FIG. 8 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0137] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0138] Acquire a symbol sequence sample, wherein the symbol sequence sample includes a plurality of symbols arranged in sequence according to a symbol reception order;

[0139] Using a sliding window to perform sliding window processing on the multiple symbols according to the symbol receiving order, for the multiple target symbols currently in the sliding window, inputting the multiple target symbols into a feedforward equalizer to obtain a feedforward output;

[0140] For each feedforward output after the Lth feedforward output, obtain the L feedforward outputs before the current feedforward output, and input the L feedforward outputs into the feedback equalizer to obtain a feedback output;

[0141] Determining a prediction sign based on the current feedforward output and the feedback output;

[0142] Based on the predicted symbol and the expected response, a prediction error is determined, and based on the prediction error, parameters in the feedforward equalizer and the feedback equalizer are adjusted until the calculated prediction error is less than a preset value.

[0143] In one embodiment, the size of the sliding window is 2K1, and the number of symbols in the symbol sequence sample is N; when the processor executes the computer program, the following steps are also implemented: extracting the 1st to 2K1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; after obtaining the corresponding feedforward output, extracting the 2nd to 2K1+1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; after obtaining the corresponding feedforward output, continuing to perform the extraction operation until the current last target symbol in the sliding window is the Nth symbol in the symbol sequence sample.

[0144] In one embodiment, the feedforward equalizer includes an input layer, a hidden layer and an output layer; the number of neurons in the input layer is the same as the number of multiple target symbols, and the number of neurons in the hidden layer is the same as the order of the channel that the equalizer needs to fit; when the processor executes the computer program, the following steps are also implemented: the multiple target symbols are respectively input into the multiple first neurons in the input layer; each first neuron in the i-th set inputs the received target symbols into the i-th order neuron in the hidden layer; the i-th set is a set composed of a preset number of target symbols before the center symbol and a preset number of target symbols after the center symbol among the multiple target symbols, each corresponding to the first neurons; each order of neurons in the hidden layer performs corresponding operations on the received target symbols, obtains the operation results and outputs them to the output layer; the output layer adds the received operation results to obtain a feedforward output.

[0145] In one embodiment, the processor further implements the following steps when executing the computer program: for the i-th order neuron, obtain the i-th tap coefficient corresponding to each received target symbol; multiply the i-th order tap coefficient corresponding to each target symbol by the corresponding target symbol, sum the multiplication results, add the sum result and the bias corresponding to the i-th order neuron, and raise the sum result to the i-th power to obtain the operation result.

[0146] In one embodiment, when the processor executes the computer program, the following steps are also implemented: obtaining the decision coefficient corresponding to the i-th feedforward output before the current feedforward output; multiplying the decision coefficient corresponding to the i-th feedforward output and the i-th feedforward output to obtain the decision result corresponding to the i-th feedforward output; adding the decision results corresponding to each of the L feedforward outputs to obtain the feedback output.

[0147] In one embodiment, the parameters of the feedforward equalizer include tap coefficients, and the parameters of the feedback equalizer include decision coefficients; when the processor executes the computer program, the following steps are also implemented: obtaining the mean square error gradient of the prediction error; for the tap coefficient or the decision coefficient, obtaining the current value of the current coefficient, multiplying the learning rate by the mean square error gradient to obtain a multiplication result; subtracting the multiplication result from the current value, and updating the current value using the subtraction result.

[0148] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0149] Acquire a symbol sequence sample, wherein the symbol sequence sample includes a plurality of symbols arranged in sequence according to a symbol reception order;

[0150] Using a sliding window to perform sliding window processing on the multiple symbols according to the symbol receiving order, for the multiple target symbols currently in the sliding window, inputting the multiple target symbols into a feedforward equalizer to obtain a feedforward output;

[0151] For each feedforward output after the Lth feedforward output, obtain the L feedforward outputs before the current feedforward output, and input the L feedforward outputs into the feedback equalizer to obtain a feedback output;

[0152] Determining a prediction sign based on the current feedforward output and the feedback output;

[0153] Based on the predicted symbol and the expected response, a prediction error is determined, and based on the prediction error, parameters in the feedforward equalizer and the feedback equalizer are adjusted until the calculated prediction error is less than a preset value.

[0154] In one embodiment, the size of the sliding window is 2K1, and the number of symbols in the symbol sequence sample is N; when the computer program is executed by the processor, the following steps are also implemented: extracting the 1st to 2K1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; after obtaining the corresponding feedforward output, extracting the 2nd to 2K1+1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; after obtaining the corresponding feedforward output, continuing to perform the extraction operation until the current last target symbol in the sliding window is the Nth symbol in the symbol sequence sample.

[0155] In one embodiment, the feedforward equalizer includes an input layer, a hidden layer and an output layer; the number of neurons in the input layer is the same as the number of multiple target symbols, and the number of neurons in the hidden layer is the same as the order of the channel that the equalizer needs to fit; when the computer program is executed by the processor, the following steps are also implemented: the multiple target symbols are respectively input into the multiple first neurons in the input layer; each first neuron in the i-th set inputs the received target symbols into the i-th order neuron in the hidden layer; the i-th set is a set composed of a preset number of target symbols before the center symbol and a preset number of target symbols after the center symbol among the multiple target symbols, each corresponding to the first neurons; each order of neurons in the hidden layer performs corresponding operations on the received target symbols, obtains the operation results and outputs them to the output layer; the output layer adds the received operation results to obtain a feedforward output.

[0156] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: for the i-th order neuron, the i-th tap coefficient corresponding to each received target symbol is obtained; the i-th order tap coefficient corresponding to each target symbol is multiplied by the corresponding target symbol, the multiplication results are summed, the summed result and the bias corresponding to the i-th order neuron are added, the i-th power of the summed result is obtained, and the operation result is obtained.

[0157] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: obtaining the decision coefficient corresponding to the i-th feedforward output before the current feedforward output; multiplying the decision coefficient corresponding to the i-th feedforward output and the i-th feedforward output to obtain the decision result corresponding to the i-th feedforward output; adding the decision results corresponding to each of the L feedforward outputs to obtain the feedback output.

[0158] In one embodiment, the parameters of the feedforward equalizer include tap coefficients, and the parameters of the feedback equalizer include decision coefficients; when the computer program is executed by the processor, the following steps are also implemented: obtaining the mean square error gradient of the prediction error; for the tap coefficient or the decision coefficient, obtaining the current value of the current coefficient, multiplying the learning rate by the mean square error gradient to obtain a multiplication result; subtracting the multiplication result from the current value, and updating the current value using the subtraction result.

[0159] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0160] Acquire a symbol sequence sample, wherein the symbol sequence sample includes a plurality of symbols arranged in sequence according to a symbol reception order;

[0161] Using a sliding window to perform sliding window processing on the multiple symbols according to the symbol receiving order, for the multiple target symbols currently in the sliding window, inputting the multiple target symbols into a feedforward equalizer to obtain a feedforward output;

[0162] For each feedforward output after the Lth feedforward output, obtain the L feedforward outputs before the current feedforward output, and input the L feedforward outputs into the feedback equalizer to obtain a feedback output;

[0163] Determining a prediction sign based on the current feedforward output and the feedback output;

[0164] Based on the predicted symbol and the expected response, a prediction error is determined, and based on the prediction error, parameters in the feedforward equalizer and the feedback equalizer are adjusted until the calculated prediction error is less than a preset value.

[0165] In one embodiment, the size of the sliding window is 2K1, and the number of symbols in the symbol sequence sample is N; when the computer program is executed by the processor, the following steps are also implemented: extracting the 1st to 2K1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; after obtaining the corresponding feedforward output, extracting the 2nd to 2K1+1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; after obtaining the corresponding feedforward output, continuing to perform the extraction operation until the current last target symbol in the sliding window is the Nth symbol in the symbol sequence sample.

[0166] In one embodiment, the feedforward equalizer includes an input layer, a hidden layer and an output layer; the number of neurons in the input layer is the same as the number of multiple target symbols, and the number of neurons in the hidden layer is the same as the order of the channel that the equalizer needs to fit; when the computer program is executed by the processor, the following steps are also implemented: the multiple target symbols are respectively input into the multiple first neurons in the input layer; each first neuron in the i-th set inputs the received target symbols into the i-th order neuron in the hidden layer; the i-th set is a set composed of a preset number of target symbols before the center symbol and a preset number of target symbols after the center symbol among the multiple target symbols, each corresponding to the first neurons; each order of neurons in the hidden layer performs corresponding operations on the received target symbols, obtains the operation results and outputs them to the output layer; the output layer adds the received operation results to obtain a feedforward output.

[0167] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: for the i-th order neuron, the i-th tap coefficient corresponding to each received target symbol is obtained; the i-th order tap coefficient corresponding to each target symbol is multiplied by the corresponding target symbol, the multiplication results are summed, the summed result and the bias corresponding to the i-th order neuron are added, the i-th power of the summed result is obtained, and the operation result is obtained.

[0168] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: obtaining the decision coefficient corresponding to the i-th feedforward output before the current feedforward output; multiplying the decision coefficient corresponding to the i-th feedforward output and the i-th feedforward output to obtain the decision result corresponding to the i-th feedforward output; adding the decision results corresponding to each of the L feedforward outputs to obtain the feedback output.

[0169] In one embodiment, the parameters of the feedforward equalizer include tap coefficients, and the parameters of the feedback equalizer include decision coefficients; when the computer program is executed by the processor, the following steps are also implemented: obtaining the mean square error gradient of the prediction error; for the tap coefficient or the decision coefficient, obtaining the current value of the current coefficient, multiplying the learning rate by the mean square error gradient to obtain a multiplication result; subtracting the multiplication result from the current value, and updating the current value using the subtraction result.

[0170] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0171] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0172] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for constructing a channel equalizer, characterized in that: The method comprises: Acquire a symbol sequence sample, wherein the symbol sequence sample includes a plurality of symbols arranged in sequence according to a symbol reception order; Using a sliding window to perform sliding window processing on the multiple symbols according to the symbol receiving order, for the multiple target symbols currently in the sliding window, inputting the multiple target symbols into a feedforward equalizer to obtain a feedforward output; For each feedforward output after the Lth feedforward output, obtain the L feedforward outputs before the current feedforward output, and input the L feedforward outputs into the feedback equalizer to obtain a feedback output; Determining a prediction sign based on the current feedforward output and the feedback output; Based on the predicted symbol and the expected response, a prediction error is determined, and based on the prediction error, parameters in the feedforward equalizer and the feedback equalizer are adjusted until the calculated prediction error is less than a preset value.

2. The method according to claim 1, characterized in that The size of the sliding window is 2K1, and the number of symbols in the symbol sequence sample is N; the using the sliding window to perform sliding window processing on the multiple symbols according to the symbol reception order includes: Extracting the 1st to 2K1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; After obtaining the corresponding feedforward output, extracting the 2nd to 2K1+1th symbols from the symbol sequence samples as the current multiple target symbols in the sliding window; After the corresponding feedforward output is obtained, the extraction operation is continued until the current last target symbol in the sliding window is the Nth symbol in the symbol sequence sample.

3. The method according to claim 1, characterized in that: The feedforward equalizer comprises an input layer, a hidden layer and an output layer; the number of neurons in the input layer is the same as the number of multiple target symbols, and the number of neurons in the hidden layer is the same as the order of the channel that the equalizer needs to fit; The step of inputting the plurality of target symbols into a feedforward equalizer to obtain a feedforward output comprises: Inputting the plurality of target symbols into the plurality of first neurons in the input layer respectively; Each first neuron in the i-th set inputs the received target symbols into the i-th order neuron in the hidden layer; the i-th set is a set consisting of the first neurons corresponding to a preset number of target symbols before the central symbol and a preset number of target symbols after the central symbol among the multiple target symbols; Each order of neurons in the hidden layer performs corresponding operations on the received target symbols, obtains the operation results and outputs them to the output layer; The output layer adds the received operation results to obtain a feedforward output.

4. The method according to claim 3, characterized in that Each order of neurons in the hidden layer performs corresponding operations on the received target symbols, including: For the i-th order neuron, obtain the i-th tap coefficient corresponding to each received target symbol; The i-th order tap coefficient corresponding to each target symbol is multiplied by the corresponding target symbol, the multiplication results are summed, the summed result and the bias corresponding to the i-th order neuron are added, the i-th power of the summed result is obtained, and the operation result is obtained.

5. The method according to claim 1, characterized in that The step of inputting the L times feedforward output to a feedback equalizer to obtain a feedback output comprises: Get the decision coefficient corresponding to the i-th feedforward output before the current feedforward output; Multiply the decision coefficient corresponding to the i-th feedforward output by the i-th feedforward output to obtain the decision result corresponding to the i-th feedforward output; The decision results corresponding to the L feedforward outputs are added together to obtain the feedback output.

6. The method according to any one of claims 1 to 5, characterized in that: The parameters of the feedforward equalizer include tap coefficients, and the parameters of the feedback equalizer include decision coefficients; and adjusting the parameters in the feedforward equalizer and the feedback equalizer based on the prediction error includes: Obtaining the mean square error gradient of the prediction error; For the tap coefficient or the decision coefficient, obtain the current value of the current coefficient, use the learning rate to multiply the mean square error gradient to obtain a multiplication result; use the current value to subtract the multiplication result, and use the subtraction result to update the current value.

7. A channel equalizer construction device, characterized in that: The device comprises: An acquisition module, used for acquiring a symbol sequence sample, wherein the symbol sequence sample comprises a plurality of symbols arranged in sequence according to a symbol reception order; A sliding window processing module, configured to perform sliding window processing on the multiple symbols according to the symbol receiving order using a sliding window, and for multiple target symbols currently in the sliding window, input the multiple target symbols into a feedforward equalizer to obtain a feedforward output; The acquisition module is further used to acquire L feedforward outputs before the current feedforward output for each feedforward output after the Lth feedforward output, input the L feedforward outputs to the feedback equalizer to obtain feedback output; and determine the prediction symbol based on the current feedforward output and the feedback output; A parameter adjustment module is used to determine a prediction error based on the predicted symbol and the expected response, and adjust parameters in the feedforward equalizer and the feedback equalizer based on the prediction error until the calculated prediction error is less than a preset value.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images