A method, system, product, device and storage medium for generating a received signal

Abstract

The present application discloses a method, system, product, device and storage medium for generating a received signal, and relates to the field of wireless communication technology, including: inputting a transmitted signal into a preset signal generation model, obtaining a received signal output by the preset signal generation model, wherein a preset generative adversarial network trains the initial signal generation model based on the degree of difference between a field-collected signal and an output result of the initial signal generation model, and obtains the preset signal generation model, and the field-collected signal includes a measured transmitted field signal and a received field signal. The present application trains the initial signal generation model based on a preset generative adversarial network, so that the obtained preset signal generation model fully considers the channel information of the signal in the measured scenario, improves the accuracy of the received signal output by the preset signal generation model, and thereby improves the test and evaluation accuracy of the functional performance of the wireless communication device.

Description

Technical Field

[0001] The present application relates to the field of wireless communication technology, and in particular to a method, system, product, device and storage medium for generating a received signal. Background Art

[0002] In order to ensure that the functions of wireless communication equipment meet the design requirements, it is necessary to test and evaluate the functional performance of the wireless communication equipment, that is, to use the wireless communication equipment to send signals and analyze the differences between the collected received signals and the sent signals, so as to evaluate the functional performance of the wireless communication equipment. However, due to the long transmission distance of wireless communication equipment, the cost of collecting received signals is high, so the existing received signals are mostly generated by analog methods.

[0003] The existing simulation method is to use channel simulation tools (such as Matlab) to model the transmission channel of the signal to add channel information to the transmitted signal to simulate the received signal. However, since the above channel information is set by the tester based on experience, the simulated channel information is quite different from the channel information of the real transmission channel, which reduces the accuracy of the received signal and further reduces the test and evaluation accuracy of the functional performance of the wireless communication device. Summary of the invention

[0004] In view of the above problems, the present application provides a method, system, product, device and storage medium for generating a received signal, so as to achieve the purpose of improving the accuracy of generating a received signal and improving the accuracy of testing and evaluating the functional performance of a wireless communication device. The specific scheme is as follows:

[0005] The first aspect of the present application provides a method for generating a received signal, comprising:

[0006] A sending signal is input into a preset signal generation model to obtain a receiving signal output by the preset signal generation model, wherein the preset signal generation model is a model trained by a preset generative adversarial network, and the preset generative adversarial network trains the initial signal generation model based on the degree of difference between the field collection signal and the output result of the initial signal generation model to obtain the preset signal generation model, and the field collection signal includes a measured sending field signal and a receiving field signal.

[0007] In a possible implementation, the preset generative adversarial network includes the preset signal generation model, the preset signal restoration model, the preset first discrimination model and the preset second discrimination model. The preset generative adversarial network trains the initial signal generation model based on the difference between the field collected signal and the output result of the initial signal generation model to obtain the preset signal generation model, including:

[0008] Obtaining a plurality of the field acquisition signals, wherein the sending field signal and the receiving field signal in the field acquisition signals have a corresponding relationship, and the channel information of each of the field acquisition signals is different;

[0009] Fix the model parameters of the initial signal generation model and the initial signal restoration model, input each of the sending field signals into the initial signal generation model respectively, obtain the initial test receiving signal corresponding to each of the sending field signals, input each of the receiving field signals into the initial signal restoration model respectively, obtain the initial test sending signal corresponding to each of the receiving field signals; based on the correspondence between the sending field signals and the receiving field signals, obtain a plurality of initial sending test data groups and a plurality of initial receiving test data groups, the initial sending test data group includes one of the sending field signals and the receiving field signals having a corresponding relationship with the receiving field signals and an initial test sending signal, and the initial receiving test data group includes one of the receiving field signals and the initial test receiving signal having a corresponding relationship with the sending field signals;

[0010] The initial first distinguishing model is trained using each of the initial sending test data groups, and the initial second distinguishing model is trained using each of the initial receiving test data groups to obtain the preset first distinguishing model and the preset second distinguishing model; based on each of the on-site collected signals, the preset first distinguishing model and the preset second distinguishing model, the initial signal generation model and the initial signal restoration model are trained to obtain the preset signal generation model and the preset signal restoration model.

[0011] In a possible implementation, the initial signal generation model and the initial signal restoration model are trained based on each of the field collected signals, the preset first distinction model, and the preset second distinction model to obtain the preset signal generation model and the preset signal restoration model, including:

[0012] Input the transmitted field signal into the initial signal generation model to obtain a test received signal, and based on the correspondence between the transmitted field signal and the received field signal, obtain a received test data group, wherein the received test data group includes the test received signal and the received field signal respectively corresponding to the transmitted field signal; input the test received signal into the initial signal restoration model to obtain a restored transmitted signal; input the received test data group into the preset first distinction model to obtain a first distinction result; and adjust the parameters of the initial signal generation model based on the difference between the restored transmitted signal and the transmitted field signal, and the first distinction result;

[0013] Input the received field signal into the initial signal restoration model to obtain a test transmission signal, and based on the correspondence between the received field signal and the transmission field signal, obtain a transmission test data group, wherein the transmission test data group includes the test transmission signal and the transmission field signal respectively corresponding to the received field signal; input the test transmission signal into the initial signal generation model to obtain a simulated received signal; input the transmission test data group into the preset second distinction model to obtain a second distinction result; and adjust the parameters of the initial signal restoration model based on the difference between the simulated received signal and the received field signal, and the second distinction result;

[0014] When the preset termination training condition is met, the training of the initial signal generation model and the initial signal restoration model is completed to obtain the preset signal generation model and the preset signal restoration model.

[0015] In a possible implementation, using each of the initial sending test data groups to train an initial first distinguishing model, using each of the initial receiving test data groups to train an initial second distinguishing model, and obtaining the preset first distinguishing model and the preset second distinguishing model, includes:

[0016] After each training of the initial first distinguishing model and the initial second distinguishing model, the formula:

[0017]

[0018] ,

[0019] Calculate the maximum discrimination After a preset number of trainings, the maximum discrimination with the largest value is selected. The corresponding model parameters are used to set the parameters of the initial first differentiation model and the initial second differentiation model to obtain the preset first differentiation model and the preset second differentiation model, wherein the is the maximum differentiation result output by the initial second differentiation model during the training process, is the maximum differentiation result output by the initial first differentiation model during the training process, and the D Y Characterize the initial first differentiation model, the D X Characterize the initial second differentiation model, the G(x) is the simulated received signal, the F(y) is the restored transmitted signal, the represents the sending scene signal, x is any one of the sending scene signals, the Characterizes that the received field signal y is any one of the received field signals, the D X(x) is the probability value of the transmitted scene signal x output by the initial second discrimination model being accurate, and the D Y (y) is the probability value of the received scene signal y output by the initial first discrimination model being accurate, and the D X (F(y)) is the probability value of the restored transmission signal F(y) output by the initial second discrimination model being accurate, and the D Y (G(x)) is the probability value that the simulated received signal G(x) output by the initial first discrimination model is accurate.

[0020] In a possible implementation, when a preset termination training condition is met, completing the training of the initial signal generation model and the initial signal restoration model to obtain the preset signal generation model and the preset signal restoration model includes:

[0021] After adjusting the parameters of the initial signal generation model and the initial signal restoration model in sequence, the formula is:

[0022]

[0023] ,

[0024] Obtain the minimum value of the difference between the simulated received signal output by the initial signal generation model and the received field signal , wherein the Characterizes that the transmission site signal x is any one of the transmission site signals, Characterizes that the received field signal y is any one of the received field signals, the D Y (G(x)) is the first distinction result output by the preset first distinction model according to the receiving test data group, and the content of the first distinction result is the probability that the simulated receiving signal G(x) output by the initial signal generation model is accurate. X (F(y)) is the second distinction result output by the preset second distinction model according to the transmission test data group, the content of the second distinction result is the probability that the restored transmission signal F(y) output by the initial signal restoration model is accurate, and G represents the initial signal generation model;

[0025] After a preset number of trainings, the model parameters corresponding to the difference with the smallest value are selected to perform parameter adjustment operations on the initial signal generation model and the initial signal restoration model. After the parameter adjustment operation is completed, it is determined whether the output value of the preset loss function is not greater than the preset threshold. If so, the training of the initial signal generation model and the initial signal restoration model is completed to obtain the preset signal generation model and the preset signal restoration model.

[0026] In a possible implementation, the determining whether the output value of the preset loss function is not greater than a preset threshold value includes:

[0027] By formula:

[0028] ,

[0029] Obtain the output value LOSS of the preset loss function, where the Loss GAN Characterize the accuracy of the preset first distinguishing model, the preset second distinguishing model and the initial signal generation model after the parameter adjustment operation, the Loss cycle It is used to characterize the difference between the input and output data contents of the initial signal generation model. identity It is used to characterize the degree of change of the parameters of the initial signal generation model when it outputs the simulated received signal. distribute It is used to characterize the degree of difference between the data distribution of each of the received field signals and the data distribution of each of the simulated received signals.

[0030] A second aspect of the present application provides a system for generating a received signal, comprising:

[0031] A signal generation module is used to input a sending signal into a preset signal generation model to obtain a receiving signal output by the preset signal generation model, wherein the preset signal generation model is a model trained by a preset generative adversarial network, and the preset generative adversarial network trains the initial signal generation model based on the degree of difference between the field acquisition signal and the output result of the initial signal generation model to obtain the preset signal generation model, and the field acquisition signal includes a measured sending field signal and a receiving field signal.

[0032] In a possible implementation, the received signal generation system further includes a model training model, and the model training module trains the initial signal generation model based on the difference between the output results of the field collected signal and the initial signal generation model by using the preset generative adversarial network to obtain the preset signal generation model, including:

[0033] The preset generative adversarial network includes the preset signal generation model, the preset signal restoration model, the preset first distinction model and the preset second distinction model,

[0034] The model training module obtains a plurality of the field acquisition signals, wherein the sending field signal and the receiving field signal in the field acquisition signals have a corresponding relationship, and the channel information of each of the field acquisition signals is different;

[0035] Fix the model parameters of the initial signal generation model and the initial signal restoration model, input each of the sending field signals into the initial signal generation model respectively, obtain the initial test receiving signal corresponding to each of the sending field signals, input each of the receiving field signals into the initial signal restoration model respectively, obtain the initial test sending signal corresponding to each of the receiving field signals; based on the correspondence between the sending field signals and the receiving field signals, obtain a plurality of initial sending test data groups and a plurality of initial receiving test data groups, the initial sending test data group includes one of the sending field signals and the receiving field signals having a corresponding relationship with the receiving field signals and an initial test sending signal, and the initial receiving test data group includes one of the receiving field signals and the initial test receiving signal having a corresponding relationship with the sending field signals;

[0036] The initial first distinguishing model is trained using each of the initial sending test data groups, and the initial second distinguishing model is trained using each of the initial receiving test data groups to obtain the preset first distinguishing model and the preset second distinguishing model; based on each of the on-site collected signals, the preset first distinguishing model and the preset second distinguishing model, the initial signal generation model and the initial signal restoration model are trained to obtain the preset signal generation model and the preset signal restoration model.

[0037] In a possible implementation, the model training model is set to: when the initial signal generation model and the initial signal restoration model are trained based on each of the field acquisition signals, the preset first distinction model, and the preset second distinction model, and the preset signal generation model and the preset signal restoration model are obtained,

[0038] Input the transmitted field signal into the initial signal generation model to obtain a test received signal, and based on the correspondence between the transmitted field signal and the received field signal, obtain a received test data group, wherein the received test data group includes the test received signal and the received field signal respectively corresponding to the transmitted field signal; input the test received signal into the initial signal restoration model to obtain a restored transmitted signal; input the received test data group into the preset first distinction model to obtain a first distinction result; and adjust the parameters of the initial signal generation model based on the difference between the restored transmitted signal and the transmitted field signal, and the first distinction result;

[0039] Input the received field signal into the initial signal restoration model to obtain a test transmission signal, and based on the correspondence between the received field signal and the transmission field signal, obtain a transmission test data group, wherein the transmission test data group includes the test transmission signal and the transmission field signal respectively corresponding to the received field signal; input the test transmission signal into the initial signal generation model to obtain a simulated received signal; input the transmission test data group into the preset second distinction model to obtain a second distinction result; and adjust the parameters of the initial signal restoration model based on the difference between the simulated received signal and the received field signal, and the second distinction result;

[0040] When the preset termination training condition is met, the training of the initial signal generation model and the initial signal restoration model is completed to obtain the preset signal generation model and the preset signal restoration model.

[0041] In a possible implementation, the model training model is set to: when the initial first distinguishing model is trained using each of the initial sending test data groups, and the initial second distinguishing model is trained using each of the initial receiving test data groups, and the preset first distinguishing model and the preset second distinguishing model are obtained:

[0042] After each training of the initial first distinguishing model and the initial second distinguishing model, the formula:

[0043]

[0044] ,

[0045] Calculate the maximum discrimination After a preset number of trainings, the maximum discrimination with the largest value is selected. The corresponding model parameters are used to set the parameters of the initial first differentiation model and the initial second differentiation model to obtain the preset first differentiation model and the preset second differentiation model, wherein the is the maximum differentiation result output by the initial second differentiation model during the training process, is the maximum differentiation result output by the initial first differentiation model during the training process, and the D Y Characterize the initial first differentiation model, the D X Characterize the initial second differentiation model, the G(x) is the simulated received signal, the F(y) is the restored transmitted signal, the represents the sending scene signal, x is any one of the sending scene signals, the Characterizes that the received field signal y is any one of the received field signals, the D X (x) is the probability value of the transmitted scene signal x output by the initial second discrimination model being accurate, and the D Y (y) is the probability value of the received scene signal y output by the initial first discrimination model being accurate, and the D X (F(y)) is the probability value of the restored transmission signal F(y) output by the initial second discrimination model being accurate, and the D Y (G(x)) is the probability value that the simulated received signal G(x) output by the initial first discrimination model is accurate.

[0046] In a possible implementation, the model training model completes the training of the initial signal generation model and the initial signal restoration model when the preset termination training condition is met, and is set to:

[0047] After adjusting the parameters of the initial signal generation model and the initial signal restoration model in sequence, the formula is:

[0048]

[0049] ,

[0050] Obtain the minimum value of the difference between the simulated received signal output by the initial signal generation model and the received field signal , wherein the Characterizes that the transmission site signal x is any one of the transmission site signals, Characterizes that the received field signal y is any one of the received field signals, the D Y (G(x)) is the first distinction result output by the preset first distinction model according to the receiving test data group, and the content of the first distinction result is the probability that the simulated receiving signal G(x) output by the initial signal generation model is accurate.X (F(y)) is the second distinction result output by the preset second distinction model according to the transmission test data group, the content of the second distinction result is the probability that the restored transmission signal F(y) output by the initial signal restoration model is accurate, and G represents the initial signal generation model;

[0051] After a preset number of trainings, the model parameters corresponding to the difference with the smallest value are selected to perform parameter adjustment operations on the initial signal generation model and the initial signal restoration model. After the parameter adjustment operation is completed, it is determined whether the output value of the preset loss function is not greater than the preset threshold. If so, the training of the initial signal generation model and the initial signal restoration model is completed to obtain the preset signal generation model and the preset signal restoration model.

[0052] In a possible implementation, the model training model is set to:

[0053] By formula:

[0054] ,

[0055] Obtain the output value LOSS of the preset loss function, where the Loss GAN Characterize the accuracy of the preset first distinguishing model, the preset second distinguishing model and the initial signal generation model after the parameter adjustment operation, the Loss cycle It is used to characterize the difference between the input and output data contents of the initial signal generation model. identity It is used to characterize the degree of change of the parameters of the initial signal generation model when it outputs the simulated received signal. distribute It is used to characterize the degree of difference between the data distribution of each of the received field signals and the data distribution of each of the simulated received signals.

[0056] The third aspect of the present application provides a computer program product, including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements the method for generating a received signal of the first aspect or any implementation of the first aspect.

[0057] A fourth aspect of the present application provides an electronic device, comprising at least one processor and a memory connected to the processor, wherein:

[0058] The memory is used to store computer programs;

[0059] The processor is used to execute the computer program so that the electronic device can implement the method for generating a received signal according to the first aspect or any implementation manner of the first aspect.

[0060] A fifth aspect of the present application provides a computer storage medium, which carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can generate a receiving signal according to the first aspect or any implementation method of the first aspect.

[0061] By means of the above technical scheme, the present application provides a method, system, product, device and storage medium for generating a received signal. By configuring and utilizing a preset generative adversarial network, the initial signal generation model is trained based on the degree of difference between the output results of the field collection signal and the initial signal generation model. Since the field collection signal includes the measured sending field signal and the receiving field signal, the obtained preset signal generation model fully considers the channel information of the signal in the actual measurement scenario, thereby improving the accuracy of the received signal output by the preset signal generation model, thereby improving the test and evaluation accuracy of the functional performance of the wireless communication device. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the accompanying drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale.

[0063] Figure 1 A flowchart of a method for generating a received signal provided by the present application;

[0064] Figure 2 A schematic diagram of a process for obtaining a received signal provided by the present application;

[0065] Figure 3 A schematic diagram of the data transmission process of each preset model in a preset generative adversarial network provided in this application;

[0066] Figure 4 A schematic diagram of the structure of a deep classifiable convolutional layer provided for this application;

[0067] Figure 5 A schematic diagram of the training process of a preset signal generation model and a preset signal restoration model provided in this application;

[0068] Figure 6 A schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION

[0069] The following describes the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. The terms used in the implementation method section of the present application are only used to explain the specific embodiments of the present application, and are not intended to limit the present application.

[0070] The embodiments of the present application are described below in conjunction with the accompanying drawings. Those skilled in the art will appreciate that, with the development of technology and the emergence of new scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0071] The terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and need not be used to describe a specific order or sequential order. It should be understood that the terms used in this way can be interchangeable under appropriate circumstances, which is only to describe the distinction mode adopted by the objects of the same attributes when describing in the embodiments of the present application. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, so that the process, method, system, product or equipment comprising a series of units need not be limited to those units, but may include other units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0072] The first aspect of the present application provides a method for generating a received signal, such as Figure 1 As shown, the method for generating the received signal includes:

[0073] S101. Input a sending signal into a preset signal generation model to obtain a receiving signal output by the preset signal generation model, wherein the preset signal generation model is a model trained by a preset generative adversarial network. The preset generative adversarial network trains the initial signal generation model based on the degree of difference between the field collection signal and the output result of the initial signal generation model to obtain the preset signal generation model. The field collection signal includes a measured sending field signal and a receiving field signal.

[0074] It should be noted that in actual application scenarios, the sending signal is input into the preset signal generation model to obtain the flow chart of the receiving signal as shown in Figure 2 As shown. The above-mentioned transmission signal may be a signal (Data frame) generated by a wireless communication device, and the above-mentioned reception signal is a signal received by a wireless communication device that has a communication relationship with the wireless communication device that sends the transmission signal. The above-mentioned transmission signal and reception signal are both composed of a frame header, a data part and a frame tail.

[0075] It should be noted that, in actual application scenarios, the above-mentioned field acquisition signals can be sending field signals and receiving field signals extracted from measured data. Since the sending field signal is affected by the transmission distance and the transmission channel after the wireless communication device sends the sending field signal, the actually received receiving field signal will have channel information such as delay duration, attenuation, and multipath effect compared to the sending field signal. However, this type of information is relatively complex, making it impossible for existing methods based on artificial modeling and settings to accurately simulate the real signal transmission environment. However, the present application configures a preset generative adversarial network based on the degree of difference between the output results of the field acquisition signal and the initial signal generation model, trains the initial signal generation model, and obtains a preset signal generation model, so that the obtained preset signal generation model fully considers the channel information of the signal in the measured scenario, thereby improving the accuracy of the received signal output by the preset signal generation model.

[0076] It should be noted that, in actual application scenarios, the above-mentioned preset generative adversarial network is a pre-configured generative adversarial network (GAN). Generative adversarial networks usually include a generative model and a discriminative model, wherein the generative model is used to capture data distribution, and the discriminative model is used to distinguish whether the input is real data or generated sample data. The generative adversarial network determines the output result of the generative model by configuring the discriminative model, and adjusts and trains the generative model according to the determination result, so that the output result of the generative model is close to the real data. Therefore, the present application trains the initial signal generation model based on the difference between the on-site collected signal and the output result of the initial signal generation model by configuring the preset generative adversarial network, thereby improving the closeness of the received signal output by the preset signal generation model to the received signal in the real scene, and improving the accuracy of the received signal output by the preset signal generation model, thereby improving the test and evaluation accuracy of the functional performance of the wireless communication device.

[0077] The present application configures and utilizes a preset generative adversarial network to train the initial signal generation model based on the degree of difference between the field collection signal and the output result of the initial signal generation model. Since the field collection signal includes the measured sending field signal and the receiving field signal, the obtained preset signal generation model fully considers the channel information of the signal in the actual measurement scenario, improves the accuracy of the received signal output by the preset signal generation model, and thereby improves the test and evaluation accuracy of the functional performance of the wireless communication device.

[0078] In a possible implementation, the preset generative adversarial network includes a preset signal generation model, a preset signal restoration model, a preset first discrimination model, and a preset second discrimination model. The preset generative adversarial network trains the initial signal generation model based on the difference between the on-site collected signal and the output result of the initial signal generation model to obtain the preset signal generation model, including:

[0079] A plurality of field acquisition signals are obtained, wherein a sending field signal and a receiving field signal in the field acquisition signals have a corresponding relationship, and channel information of each field acquisition signal is different;

[0080] Fix the model parameters of the initial signal generation model and the initial signal restoration model, input each sending field signal into the initial signal generation model respectively, obtain the initial test receiving signal corresponding to each sending field signal, input each receiving field signal into the initial signal restoration model respectively, obtain the initial test sending signal corresponding to each receiving field signal; based on the corresponding relationship between the sending field signal and the receiving field signal, obtain multiple initial sending test data groups and multiple initial receiving test data groups, the initial sending test data group includes a sending field signal having a corresponding relationship with the receiving field signal and an initial test sending signal, and the initial receiving test data group includes a receiving field signal having a corresponding relationship with the sending field signal and an initial test receiving signal;

[0081] The initial first distinguishing model is trained using each initial sending test data group, and the initial second distinguishing model is trained using each initial receiving test data group to obtain a preset first distinguishing model and a preset second distinguishing model; based on each on-site collected signal, the preset first distinguishing model and the preset second distinguishing model, the initial signal generation model and the initial signal restoration model are trained to obtain a preset signal generation model and a preset signal restoration model.

[0082] It should be noted that, in actual application scenarios, each field collection signal may be a plurality of groups of field collection signals collected and measured under different transmission distances and different transmission channel environments.

[0083] It should be noted that in actual application scenarios, due to the high cost of collecting measured data, in order to reduce training costs, this application configures the above-mentioned preset generative adversarial network including a preset signal generation model, a preset signal restoration model, a preset first differentiation model and a preset second differentiation model. During the training process, the initial signal restoration model is used to generate a test transmission signal according to the received field signal, and the test transmission signal is used as the training data of the initial signal generation model, thereby increasing the amount of data used for training the initial signal generation model and reducing the training cost. At the same time, this application also configures a preset second differentiation model to adjust the parameters of the initial signal restoration model during the training process, thereby improving the accuracy of the test transmission signal output by the initial signal restoration model.

[0084] In a possible implementation, the data transmission process diagram of each preset model in the above preset generative adversarial network is as follows: Figure 3 As shown:

[0085] The received signal is input into the preset signal restoration model to obtain the restored transmission signal output by the preset signal restoration model. When it is necessary to optimize the parameters of the preset signal generation model, the restored transmission signal and the real transmission signal are input into the preset first distinguishing model, so that the preset first distinguishing model optimizes the parameters of the preset signal generation model according to the difference between the restored transmission signal and the real transmission signal.

[0086] The transmission signal is input into the preset signal generation model to obtain a simulated reception signal output by the preset signal generation model. When it is necessary to optimize the parameters of the preset signal restoration model, the simulated reception signal and the real reception signal are input into the preset second differentiation model, so that the preset second differentiation model optimizes the parameters of the preset signal restoration model according to the difference between the simulated reception signal and the real transmission signal. The above-mentioned reception signal can be a real reception signal, or it can be a simulated reception signal generated by the preset signal generation model according to the real transmission signal. The above-mentioned transmission signal can be a real transmission signal, or it can be a restored transmission signal generated by the preset signal restoration model according to the real transmission signal.

[0087] It should be noted that, in actual application scenarios, since the difference between sending a live signal and receiving a live signal is only the channel information in the received live signal, the model structures of the initial signal generation model and the initial signal restoration model are the same, and only the input and output of the two models need to be defined respectively. The model structures of the initial signal generation model and the initial signal restoration model are shown in Table 1 below.

[0088] Table 1

[0089]

[0090] It should be noted that, in actual application scenarios, since the above-mentioned initial first differentiation model and the above-mentioned initial second differentiation model are both used to analyze the difference between input data. Therefore, the model structures of the above-mentioned initial first differentiation model and the above-mentioned initial second differentiation model are the same, and it is only necessary to define the input and output of the quantity model respectively. The model structures of the above-mentioned initial first differentiation model and the above-mentioned initial second differentiation model are shown in Table 2 below.

[0091] It should be noted that in actual application scenarios, the structural diagram of the above-mentioned deep classifiable convolutional layer can be as follows: Figure 4As shown in the figure, the input is processed in sequence by a batch normalization layer (BatchNorm), an activation layer (Relu), a depth-wise separable convolution layer (Conv2DSeparable), a 2D convolution layer (Conv2D 1×1), a batch normalization layer (BatchNorm), an activation layer (Relu), a depth-wise separable convolution layer (Conv2DSeparable), and a 2D convolution layer (Conv2D 1×1) to obtain the output.

[0092] Table 2

[0093]

[0094] In a possible implementation, based on each field collected signal, the preset first distinguishing model and the preset second distinguishing model, the initial signal generation model and the initial signal restoration model are trained to obtain the preset signal generation model and the preset signal restoration model, including:

[0095] Input the transmitted field signal into the initial signal generation model to obtain a test received signal, and based on the corresponding relationship between the transmitted field signal and the received field signal, obtain a received test data group, wherein the received test data group includes a test received signal and a received field signal respectively corresponding to the transmitted field signal; input the test received signal into the initial signal restoration model to obtain a restored transmitted signal; input the received test data group into a preset first differentiation model to obtain a first differentiation result; and adjust the parameters of the initial signal generation model based on the difference between the restored transmitted signal and the transmitted field signal, and the first differentiation result;

[0096] Input the received field signal into the initial signal restoration model to obtain a test transmission signal, and based on the corresponding relationship between the received field signal and the transmission field signal, obtain a transmission test data group, the transmission test data group includes a test transmission signal and a transmission field signal respectively corresponding to the received field signal; input the test transmission signal into the initial signal generation model to obtain a simulated received signal; input the transmission test data group into a preset second distinction model to obtain a second distinction result; based on the difference between the simulated received signal and the received field signal, and the second distinction result, adjust the parameters of the initial signal restoration model;

[0097] When the preset termination training conditions are met, the training of the initial signal generation model and the initial signal restoration model is completed to obtain the preset signal generation model and the preset signal restoration model.

[0098] In a possible implementation, the test received signal may be a restored sent signal generated by the initial signal restoration model based on the received field signal in addition to being a signal generated by the initial signal generation model based on the sent field signal. Similarly, the test sent signal may be a simulated received signal generated by the initial signal generation model based on the sent field signal in addition to being a signal generated by the initial signal restoration model based on the received field signal.

[0099] In a possible implementation, using each initial sending test data group to train an initial first distinguishing model, using each initial receiving test data group to train an initial second distinguishing model, and obtaining a preset first distinguishing model and a preset second distinguishing model, includes:

[0100] After each training of the initial first discrimination model and the initial second discrimination model, the formula is:

[0101]

[0102] ,

[0103] Calculate the maximum discrimination After a preset number of trainings, the maximum discrimination value with the largest value is selected. The corresponding model parameters are used to set the parameters of the initial first differentiation model and the initial second differentiation model to obtain the preset first differentiation model and the preset second differentiation model, wherein: is the maximum discrimination result output by the initial second discrimination model during the training process, is the maximum discrimination result output by the initial first discrimination model during the training process, D Y Characterizing the initial first discrimination model, D X Characterize the initial second differentiation model, G(x) is the simulated received signal, F(y) is the restored transmitted signal, represents the transmitted field signal, x is any one of the transmitted field signals, Characterizes that the received field signal y is any one of the received field signals, D X (x) is the probability value of the transmission scene signal x output by the initial second discrimination model being accurate, D Y (y) is the probability value of the received scene signal y output by the initial first discrimination model being accurate, D X (F(y)) is the probability value that the restored transmission signal F(y) output by the initial second discrimination model is accurate, D Y (G(x)) is the probability value that the simulated received signal G(x) output by the initial first discrimination model is accurate.

[0104] In a possible implementation, when a preset termination training condition is met, completing the training of the initial signal generation model and the initial signal restoration model to obtain the preset signal generation model and the preset signal restoration model includes:

[0105] After adjusting the parameters of the initial signal generation model and the initial signal restoration model in sequence, the formula is:

[0106]

[0107] ,

[0108] Obtain the minimum value of the difference between the simulated received signal output by the initial signal generation model and the received on-site signal ,in, The characterization transmission scene signal x is any one of the transmission scene signals, Characterizes that the received field signal y is any one of the received field signals, D Y (G(x)) is the first distinction result output by the preset first distinction model according to the received test data set. The content of the first distinction result is the probability that the simulated received signal G(x) output by the initial signal generation model is accurate. D X (F(y)) is the second distinction result output by the preset second distinction model according to the sent test data group, the content of the second distinction result is the probability that the restored sent signal F(y) output by the initial signal restoration model is accurate, and G represents the initial signal generation model;

[0109] After a preset number of trainings, the model parameters corresponding to the smallest numerical difference are selected to perform parameter adjustment operations on the initial signal generation model and the initial signal restoration model. After the parameter adjustment operation is completed, it is determined whether the output value of the preset loss function is not greater than the preset threshold. If so, the training of the initial signal generation model and the initial signal restoration model is completed to obtain the preset signal generation model and the preset signal restoration model.

[0110] In a possible implementation, determining whether an output value of a preset loss function is not greater than a preset threshold includes:

[0111] By formula:

[0112] ,

[0113] Obtain the output value LOSS of the preset loss function, where Loss GAN Characterizes the accuracy of the preset first differentiation model, the preset second differentiation model, and the initial signal generation model after parameter adjustment, Loss cycle It is used to characterize the difference between the input and output data content of the initial signal generation model.identity It is used to characterize the degree of change of the parameters of the initial signal generation model when it outputs the simulated received signal. distribute It is used to characterize the difference between the data distribution of each received field signal and the data distribution of each simulated received signal.

[0114] In order to facilitate the understanding of the present application, the training process of the above-mentioned preset signal generation model and the preset signal restoration model is described here in combination with a possible implementation of the present application:

[0115] It is assumed that both the preset first distinguishing model and the preset second distinguishing model have completed training.

[0116] like Figure 5 As shown, it is a schematic diagram of the training process of the above-mentioned preset signal generation model and the preset signal restoration model.

[0117] The sending field signal is input into the initial signal generation model to obtain a test receiving signal; based on the corresponding relationship between the sending field signal and the receiving field signal, a receiving test data group is obtained, and the receiving test data group includes the test receiving signal and the receiving field signal respectively corresponding to the sending field signal. The receiving field signal is input into the initial signal restoration model to obtain a test sending signal, and based on the corresponding relationship between the receiving field signal and the sending field signal, a sending test data group is obtained, and the sending test data group includes the test sending signal and the sending field signal respectively corresponding to the receiving field signal.

[0118] Input the test receiving signal into the initial signal restoration model to obtain the test sending signal. Input the test sending signal into the initial signal generation model to obtain the simulated receiving signal. At this time, call the first loss function Loss cycle , at least based on the transmitted live signal and the restored transmitted signal, the difference between the data content of the transmitted live signal and the simulated received signal is evaluated. The second loss function Loss is called identity , at least based on the transmitted live signal and the restored transmitted signal, the degree of change of the initial signal generation model's own parameters when outputting the simulated received signal is evaluated.

[0119] The above first loss function formula is:

[0120] , where x is the transmitted signal, G(x) is the simulated received signal generated by the initial signal generation model according to the transmitted signal X, and F(G(x)) is the restored transmitted signal output by the initial signal restoration model using the simulated received signal generated by the initial signal generation model according to the transmitted signal x as input. y is the simulated received signal G(x), F(y) is the simulated transmitted signal generated by the initial signal restoration model according to y, and G(F(y)) is the simulated received signal generated by the initial signal generation model using the restored transmitted signal generated by the initial signal restoration model according to y as input.

[0121] It should be noted that, in actual application scenarios, the training of the initial signal generation model and the initial signal restoration model can be performed simultaneously or in steps. When performed simultaneously, the input of the initial signal restoration model is the output of the initial signal generation model, and the input of the initial signal generation model is the transmitted on-site signal. When performed in steps, the input of the initial signal restoration model is the received on-site signal, and the input of the initial signal generation model is the output of the initial signal restoration model.

[0122] The above second loss function formula is:

[0123] , where x is the restored transmission signal output by the initial signal restoration model, and F(x) is the restored transmission signal regenerated by the initial signal restoration model based on the restored transmission signal. y is the simulated reception signal output by the initial signal generation model, and G(y) is the simulated reception signal regenerated by the initial signal generation model based on the simulated reception signal.

[0124] The received test data group is input into a preset first differentiation model to obtain a first differentiation result. Based on the difference between the restored transmission signal and the transmission site signal and the first differentiation result, the initial signal generation model is adjusted.

[0125] During the parameter adjustment operation, the third loss function is called to determine the parameter adjustment directions of the initial signal generation model and the initial signal restoration model. The formula of the third loss function is: Among them, D Y Characterize the preset first distinction model, D X Characterize the preset second differentiation model, G(x) is the simulated received signal, F(y) is the restored sent signal, y is the received live signal, x is the sent live signal, D Y (y) is the probability that the received scene signal output by the preset first discrimination model is accurate, D X (x) is the probability that the transmitted scene signal output by the preset second discrimination model is accurate, D Y(G(x)) is the probability that the simulated received signal output by the preset first discrimination model is accurate, D X (F(y)) is the probability that the restored transmitted signal output by the preset second distinguishing model is accurate.

[0126] After the parameter adjustment operation is completed and the number of training times is not less than the preset threshold, the fourth loss function is called, through the formula: , to obtain the degree of difference between data distributions. Std is the standard deviation function, and avg is the mean function. Finally, the loss function formula is preset: , obtain the output value, and determine whether the output value is not greater than the preset threshold. If so, end the training, if not, repeat the above training process.

[0127] A second aspect of the present application provides a system for generating a received signal, the system for generating a received signal comprising:

[0128] A signal generation module is used to input a sending signal into a preset signal generation model to obtain a receiving signal output by the preset signal generation model, wherein the preset signal generation model is a model trained by a preset generative adversarial network, and the preset generative adversarial network trains the initial signal generation model based on the degree of difference between the field collection signal and the output result of the initial signal generation model to obtain the preset signal generation model, and the field collection signal includes the measured sending field signal and the receiving field signal.

[0129] In a possible implementation, the generation system of the received signal also includes a model training model, and the model training module uses a preset generative adversarial network to train the initial signal generation model based on the difference between the on-site collected signal and the output result of the initial signal generation model to obtain the preset signal generation model, including:

[0130] The preset generative adversarial network includes a preset signal generation model, a preset signal restoration model, a preset first distinction model and a preset second distinction model.

[0131] The model training module obtains multiple field acquisition signals, wherein the sending field signal and the receiving field signal in the field acquisition signals have a corresponding relationship, and the channel information of each field acquisition signal is different;

[0132] Fix the model parameters of the initial signal generation model and the initial signal restoration model, input each sending field signal into the initial signal generation model respectively, obtain the initial test receiving signal corresponding to each sending field signal, input each receiving field signal into the initial signal restoration model respectively, obtain the initial test sending signal corresponding to each receiving field signal; based on the corresponding relationship between the sending field signal and the receiving field signal, obtain multiple initial sending test data groups and multiple initial receiving test data groups, the initial sending test data group includes a sending field signal having a corresponding relationship with the receiving field signal and an initial test sending signal, and the initial receiving test data group includes a receiving field signal having a corresponding relationship with the sending field signal and an initial test receiving signal;

[0133] The initial first distinguishing model is trained using each initial sending test data group, and the initial second distinguishing model is trained using each initial receiving test data group to obtain a preset first distinguishing model and a preset second distinguishing model; based on each on-site collected signal, the preset first distinguishing model and the preset second distinguishing model, the initial signal generation model and the initial signal restoration model are trained to obtain a preset signal generation model and a preset signal restoration model.

[0134] In a possible implementation, the model training model is set to: when the initial signal generation model and the initial signal restoration model are trained based on each field acquisition signal, the preset first distinction model and the preset second distinction model, and the preset signal generation model and the preset signal restoration model are obtained,

[0135] Input the transmitted field signal into the initial signal generation model to obtain a test received signal, and based on the corresponding relationship between the transmitted field signal and the received field signal, obtain a received test data group, wherein the received test data group includes a test received signal and a received field signal respectively corresponding to the transmitted field signal; input the test received signal into the initial signal restoration model to obtain a restored transmitted signal; input the received test data group into a preset first differentiation model to obtain a first differentiation result; and adjust the parameters of the initial signal generation model based on the difference between the restored transmitted signal and the transmitted field signal, and the first differentiation result;

[0136] Input the received field signal into the initial signal restoration model to obtain a test transmission signal, and based on the corresponding relationship between the received field signal and the transmission field signal, obtain a transmission test data group, the transmission test data group includes a test transmission signal and a transmission field signal respectively corresponding to the received field signal; input the test transmission signal into the initial signal generation model to obtain a simulated received signal; input the transmission test data group into a preset second distinction model to obtain a second distinction result; based on the difference between the simulated received signal and the received field signal, and the second distinction result, adjust the parameters of the initial signal restoration model;

[0137] When the preset termination training conditions are met, the training of the initial signal generation model and the initial signal restoration model is completed to obtain the preset signal generation model and the preset signal restoration model.

[0138] In a possible implementation, the model training model is set to: when the initial first distinguishing model is trained using each initial sending test data group, and the initial second distinguishing model is trained using each initial receiving test data group, and the preset first distinguishing model and the preset second distinguishing model are obtained:

[0139] After each training of the initial first discrimination model and the initial second discrimination model, the formula is:

[0140]

[0141] ,

[0142] Calculate the maximum discrimination After a preset number of trainings, the maximum discrimination value with the largest value is selected. The corresponding model parameters are used to set the parameters of the initial first differentiation model and the initial second differentiation model to obtain the preset first differentiation model and the preset second differentiation model, wherein: is the maximum discrimination result output by the initial second discrimination model during the training process, is the maximum discrimination result output by the initial first discrimination model during the training process, D Y Characterizing the initial first discrimination model, D X Characterize the initial second differentiation model, G(x) is the simulated received signal, F(y) is the restored transmitted signal, represents the transmitted field signal, x is any one of the transmitted field signals, Characterizes that the received field signal y is any one of the received field signals, D X (x) is the probability value of the transmission scene signal x output by the initial second discrimination model being accurate, D Y (y) is the probability value of the received scene signal y output by the initial first discrimination model being accurate, D X (F(y)) is the probability value that the restored transmission signal F(y) output by the initial second discrimination model is accurate, D Y (G(x)) is the probability value that the simulated received signal G(x) output by the initial first discrimination model is accurate.

[0143] In a possible implementation, the model training model completes the training of the initial signal generation model and the initial signal restoration model when the preset termination training condition is met, and when the preset signal generation model and the preset signal restoration model are obtained, they are set to:

[0144] After adjusting the parameters of the initial signal generation model and the initial signal restoration model in sequence, the formula is:

[0145]

[0146] ,

[0147] Obtain the minimum value of the difference between the simulated received signal output by the initial signal generation model and the received on-site signal ,in, The characterization transmission scene signal x is any one of the transmission scene signals, Characterizes that the received field signal y is any one of the received field signals, D Y (G(x)) is the first distinction result output by the preset first distinction model according to the received test data set. The content of the first distinction result is the probability that the simulated received signal G(x) output by the initial signal generation model is accurate. D X (F(y)) is the second distinction result output by the preset second distinction model according to the sent test data group, the content of the second distinction result is the probability that the restored sent signal F(y) output by the initial signal restoration model is accurate, and G represents the initial signal generation model;

[0148] After a preset number of trainings, the model parameters corresponding to the smallest numerical difference are selected to perform parameter adjustment operations on the initial signal generation model and the initial signal restoration model. After the parameter adjustment operation is completed, it is determined whether the output value of the preset loss function is not greater than the preset threshold. If so, the training of the initial signal generation model and the initial signal restoration model is completed to obtain the preset signal generation model and the preset signal restoration model.

[0149] In a possible implementation, the model training model is set to:

[0150] By formula:

[0151] ,

[0152] Obtain the output value LOSS of the preset loss function, where Loss GAN Characterizes the accuracy of the preset first differentiation model, the preset second differentiation model, and the initial signal generation model after parameter adjustment, Loss cycle It is used to characterize the difference between the input and output data content of the initial signal generation model. identity It is used to characterize the degree of change of the parameters of the initial signal generation model when it outputs the simulated received signal. distributeIt is used to characterize the difference between the data distribution of each received field signal and the data distribution of each simulated received signal.

[0153] The third aspect of the present application provides a computer program product, including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements the method for generating a received signal of the first aspect or any implementation of the first aspect.

[0154] A fourth aspect of the present application provides an electronic device, comprising at least one processor and a memory connected to the processor, wherein:

[0155] The memory is used to store computer programs;

[0156] The processor is used to execute a computer program so that the electronic device can implement the method for generating a received signal of the above-mentioned first aspect or any implementation manner of the first aspect.

[0157] The fifth aspect of the present application provides a computer storage medium, which carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can generate a receiving signal according to the first aspect or any implementation method of the first aspect.

[0158] The fourth aspect of the present application provides a structural schematic diagram of an electronic device as shown in Figure 6 The electronic devices in the embodiments of the present application may include but are not limited to fixed terminals such as mobile phones, notebook computers, PDAs (personal digital assistants), PADs (tablet computers), desktop computers, and the like. Figure 6 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0159] like Figure 6 As shown, the electronic device may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 to a random access memory (RAM) 603. When the electronic device is powered on, various programs and data required for the operation of the electronic device are also stored in the RAM 603. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0160] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a memory card, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although Figure 6 An electronic device having various devices is shown, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0161] It should also be noted that the device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the device embodiments provided by the present application, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines.

[0162] Through the description of the above implementation mode, the technicians in the field can clearly understand that the present application can be implemented by means of software plus necessary general hardware, and of course, it can also be implemented by special hardware including special integrated circuits, special CPUs, special memories, special components, etc. In general, all functions completed by computer programs can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be various, such as analog circuits, digital circuits or special circuits. However, for the present application, software program implementation is a better implementation mode in more cases. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a readable storage medium, such as a computer floppy disk, a U disk, a mobile hard disk, a ROM, a RAM, a disk or an optical disk, etc., including a number of instructions to enable a computer device (which can be a personal computer, a training device, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0163] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0164] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a training device, or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, training device, or data center. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device, a data center, etc. that includes one or more available media integrations. The available medium may be a magnetic medium, (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)), etc.

Claims

1. A method for generating a received signal, characterized in that: include: Input a transmission signal into a preset signal generation model to obtain a reception signal output by the preset signal generation model, wherein the transmission signal is a signal generated by a wireless communication device, and the reception signal is a signal received by a wireless communication device that has a communication relationship with the wireless communication device that sends the transmission signal, and the transmission signal and the reception signal are both composed of a frame header, a data part, and a frame tail. The preset signal generation model is a model trained by a preset generative adversarial network, and the preset generative adversarial network trains the initial signal generation model based on the difference between the field collection signal and the output result of the initial signal generation model to obtain the preset signal generation model. The field collection signal includes a measured transmission field signal and a reception field signal, and the preset generative adversarial network includes the preset signal generation model, a preset signal restoration model, a preset first distinction model, and a preset second distinction model. The preset generative adversarial network trains the initial signal generation model based on the difference between the field collection signal and the output result of the initial signal generation model. The process of obtaining the preset signal generation model includes: Obtaining a plurality of the field acquisition signals, wherein the sending field signal and the receiving field signal in the field acquisition signals have a corresponding relationship, and the channel information of each of the field acquisition signals is different; Fix the model parameters of the initial signal generation model and the initial signal restoration model, input each of the sending field signals into the initial signal generation model respectively, obtain the initial test receiving signal corresponding to each of the sending field signals, input each of the receiving field signals into the initial signal restoration model respectively, obtain the initial test sending signal corresponding to each of the receiving field signals; based on the correspondence between the sending field signals and the receiving field signals, obtain a plurality of initial sending test data groups and a plurality of initial receiving test data groups, the initial sending test data group includes one of the sending field signals and the receiving field signals having a corresponding relationship with the receiving field signals and an initial test sending signal, and the initial receiving test data group includes one of the receiving field signals and the initial test receiving signal having a corresponding relationship with the sending field signals; The initial first distinguishing model is trained using each of the initial sending test data groups, and the initial second distinguishing model is trained using each of the initial receiving test data groups to obtain the preset first distinguishing model and the preset second distinguishing model; based on each of the on-site collected signals, the preset first distinguishing model and the preset second distinguishing model, the initial signal generation model and the initial signal restoration model are trained to obtain the preset signal generation model and the preset signal restoration model.

2. The method for generating a received signal according to claim 1, characterized in that: The training of the initial signal generation model and the initial signal restoration model based on each of the field collected signals, the preset first distinguishing model and the preset second distinguishing model to obtain the preset signal generation model and the preset signal restoration model includes: Input the transmitted field signal into the initial signal generation model to obtain a test received signal, and based on the correspondence between the transmitted field signal and the received field signal, obtain a received test data group, wherein the received test data group includes the test received signal and the received field signal respectively corresponding to the transmitted field signal; input the test received signal into the initial signal restoration model to obtain a restored transmitted signal; input the received test data group into the preset first distinction model to obtain a first distinction result; and adjust the parameters of the initial signal generation model based on the difference between the restored transmitted signal and the transmitted field signal, and the first distinction result; Input the received field signal into the initial signal restoration model to obtain a test transmission signal, and based on the correspondence between the received field signal and the transmission field signal, obtain a transmission test data group, wherein the transmission test data group includes the test transmission signal and the transmission field signal respectively corresponding to the received field signal; input the test transmission signal into the initial signal generation model to obtain a simulated received signal; input the transmission test data group into the preset second distinction model to obtain a second distinction result; and adjust the parameters of the initial signal restoration model based on the difference between the simulated received signal and the received field signal, and the second distinction result; When the preset termination training condition is met, the training of the initial signal generation model and the initial signal restoration model is completed to obtain the preset signal generation model and the preset signal restoration model.

3. The method for generating a received signal according to claim 1, characterized in that: The method of training the initial first distinguishing model by using each of the initial sending test data groups, and training the initial second distinguishing model by using each of the initial receiving test data groups, to obtain the preset first distinguishing model and the preset second distinguishing model, comprises: After each training of the initial first distinguishing model and the initial second distinguishing model, the formula: , calculate the maximum discrimination After a preset number of trainings, the maximum discrimination with the largest value is selected. The corresponding model parameters are used to set the parameters of the initial first differentiation model and the initial second differentiation model to obtain the preset first differentiation model and the preset second differentiation model, wherein the is the maximum differentiation result output by the initial second differentiation model during the training process, is the maximum differentiation result output by the initial first differentiation model during the training process, and the D Y Characterize the initial first differentiation model, the D X Characterize the initial second differentiation model, the G(x) is the simulated received signal, the F(y) is the restored transmitted signal, the represents the sending scene signal, x is any one of the sending scene signals, the Characterizes that the received field signal y is any one of the received field signals, the D X (x) is the probability value of the transmitted scene signal x output by the initial second discrimination model being accurate, and the D Y (y) is the probability value of the received scene signal y output by the initial first discrimination model being accurate, and the D X (F(y)) is the probability value of the restored transmission signal F(y) output by the initial second discrimination model being accurate, and the D Y (G(x)) is the probability value that the simulated received signal G(x) output by the initial first discrimination model is accurate.

4. The method for generating a received signal according to claim 2, wherein: The step of completing the training of the initial signal generation model and the initial signal restoration model to obtain the preset signal generation model and the preset signal restoration model when the preset training termination condition is met includes: After adjusting the parameters of the initial signal generation model and the initial signal restoration model in sequence, the formula is: Obtain the minimum value of the difference between the simulated received signal output by the initial signal generation model and the received field signal , wherein the Characterizes that the transmission site signal x is any one of the transmission site signals, Characterizes that the received field signal y is any one of the received field signals, the D Y (G(x)) is the first distinction result output by the preset first distinction model according to the receiving test data group, and the content of the first distinction result is the probability that the simulated receiving signal G(x) output by the initial signal generation model is accurate. X (F(y)) is the second distinction result output by the preset second distinction model according to the transmission test data group, the content of the second distinction result is the probability that the restored transmission signal F(y) output by the initial signal restoration model is accurate, and G represents the initial signal generation model; After a preset number of trainings, the model parameters corresponding to the difference with the smallest value are selected to perform parameter adjustment operations on the initial signal generation model and the initial signal restoration model. After the parameter adjustment operation is completed, it is determined whether the output value of the preset loss function is not greater than the preset threshold. If so, the training of the initial signal generation model and the initial signal restoration model is completed to obtain the preset signal generation model and the preset signal restoration model.

5. The method for generating a received signal according to claim 4, characterized in that: The determining whether the output value of the preset loss function is not greater than a preset threshold value includes: By formula: Obtain the output value LOSS of the preset loss function, where the Loss GAN Characterize the accuracy of the preset first distinguishing model, the preset second distinguishing model and the initial signal generation model after the parameter adjustment operation, the Loss cycle It is used to characterize the difference between the input and output data contents of the initial signal generation model. identity It is used to characterize the degree of change of the parameters of the initial signal generation model when it outputs the simulated received signal. distribute It is used to characterize the degree of difference between the data distribution of each of the received field signals and the data distribution of each of the simulated received signals.

6. A system for generating a received signal, characterized in that: include: A signal generation module, configured to input a transmission signal into a preset signal generation model, and obtain a reception signal output by the preset signal generation model, wherein the transmission signal is a signal generated by a wireless communication device, and the reception signal is a signal received by a wireless communication device that simulates a communication relationship with the wireless communication device that sends the transmission signal, and the transmission signal and the reception signal are both composed of a frame header, a data portion, and a frame tail. The preset signal generation model is a model trained by a model training module using a preset generative adversarial network, and the model training module uses the preset generative adversarial network to train the initial signal generation model based on the difference between the on-site collection signal and the output result of the initial signal generation model to obtain the preset signal generation model. The on-site collection signal includes a measured transmission on-site signal and a reception on-site signal, and the preset generative adversarial network includes the preset signal generation model, a preset signal restoration model, a preset first discrimination model, and a preset second discrimination model. When the model training module uses the preset generative adversarial network to train the initial signal generation model based on the difference between the on-site collection signal and the output result of the initial signal generation model to obtain the preset signal generation model, the preset signal generation model is set to: The model training module obtains a plurality of the field acquisition signals, wherein the sending field signal and the receiving field signal in the field acquisition signals have a corresponding relationship, and the channel information of each of the field acquisition signals is different; the model parameters of the initial signal generation model and the initial signal restoration model are fixed, each of the sending field signals is input into the initial signal generation model respectively, and an initial test receiving signal corresponding to each of the sending field signals is obtained, and each of the receiving field signals is input into the initial signal restoration model respectively, and an initial test sending signal corresponding to each of the receiving field signals is obtained; based on the corresponding relationship between the sending field signal and the receiving field signal, a plurality of initial sending test data groups and a plurality of initial receiving test data groups are obtained, the initial sending test data group includes a sending field signal and an initial test sending signal having a corresponding relationship with the receiving field signal, and the initial receiving test data group includes a receiving field signal and an initial test receiving signal having a corresponding relationship with the sending field signal; the initial first distinguishing model is trained using each of the initial sending test data groups, and the initial second distinguishing model is trained using each of the initial receiving test data groups, and the preset first distinguishing model and the preset second distinguishing model are obtained; Based on each of the field collected signals, the preset first distinguishing model and the preset second distinguishing model, the initial signal generation model and the initial signal restoration model are trained to obtain the preset signal generation model and the preset signal restoration model.

7. A computer program product, characterized in that The method comprises computer-readable instructions, and when the computer-readable instructions are executed on an electronic device, the electronic device implements the method for generating a received signal as claimed in any one of claims 1 to 5.

8. An electronic device, characterized in that: The method comprises at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs; The processor is used to execute the computer program so that the electronic device can implement the method for generating a received signal as described in any one of claims 1 to 5.

9. A computer storage medium, characterized in that The storage medium carries one or more computer programs, and when the one or more computer programs are executed by an electronic device, the electronic device can implement the method for generating a received signal as described in any one of claims 1 to 5.

Images