Signal generation method, storage medium, electronic apparatus, and product

Abstract

Provided in the embodiments of the present disclosure are a signal generation method, a storage medium, an electronic apparatus, and a product. The method comprises: modulating data bits to generate a data symbol, and modulating pilot bits to generate a pilot symbol; superimposing the data symbol and the pilot symbol on the same time-frequency resource, so as to generate superimposed data; and acquiring a superimposed signal on the basis of the superimposed data, and sending the superimposed signal to a receiving end. By means of the embodiments of the present disclosure, the problem in the related art of low efficiency due to a pilot and data being isolated from each other on a time-frequency resource is solved.

Description

A signal generation method, storage medium, electronic device, and product.

[0001] Cross-references to related applications

[0002] This disclosure is based on and claims priority to Chinese patent application CN202510121722.0, filed on January 24, 2025, entitled “A signal generation method, storage medium, electronic device and product”, and incorporates the entire contents of that patent application by reference. Technical Field

[0003] This disclosure relates to the field of communications, and more specifically, to a signal generation method, a storage medium, an electronic device, and a product. Background Technology

[0004] In traditional wireless communication systems, channel estimation is a crucial step for achieving reliable data transmission. A typical approach is to independently allocate pilot signals across time and frequency resources. Pilot signals are inserted into the transmission frame at predetermined intervals. The receiver uses these pilot signals to estimate the current channel state and then uses the estimated channel state information to demodulate the data. In this process, pilot signals and data signals are strictly separated in both the time and frequency domains to avoid mutual interference. This leads to the occupancy of time and frequency resources for the data signal, resulting in a waste of valuable resources. Furthermore, because the allocation of pilot signals is fixed, it is difficult to flexibly adjust them according to different channel conditions, thus limiting the system's spectral efficiency. In the resource-constrained environment of modern wireless communication, this efficiency loss will significantly impact the overall system performance. Summary of the Invention

[0005] This disclosure provides a signal generation method, storage medium, electronic device, and product to at least solve the problem of low efficiency caused by the isolation of pilot signals and data in time and frequency resources in the related art.

[0006] According to one embodiment of this disclosure, a signal generation method is provided, applied at a transmitting end, comprising: modulating data bits to generate data symbols, modulating pilot bits to generate pilot symbols; superimposing the data symbols and pilot symbols on the same time-frequency resources to generate aliased data; obtaining an aliased signal based on the aliased data, and sending the aliased signal to a receiving end.

[0007] According to another embodiment of this disclosure, a signal processing method is provided, applied at a receiving end, comprising: receiving an aliased signal; converting the aliased signal into a receiving end frequency domain signal; performing energy normalization processing on the receiving end frequency domain signal to obtain a first signal; performing data integration on the first signal and a pilot signal to obtain a second signal; inputting the second signal into an artificial intelligence (AI) receiver to obtain data symbol bit information and the log-likelihood ratio (LLR) value of the data symbol bit information; and verifying the correctness of the AI ​​receiver's decoding and evaluating the transmission performance based on the LLR value.

[0008] According to yet another embodiment of this disclosure, a computer-readable storage medium is also provided, wherein a computer program is stored therein, wherein the computer program is configured to perform the steps in any of the above method embodiments when it is run.

[0009] According to yet another embodiment of this disclosure, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0010] According to yet another embodiment of this disclosure, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps in any of the above method embodiments. Attached Figure Description

[0011] Figure 1 is a hardware structure block diagram of a computer terminal for a signal generation method according to an embodiment of the present disclosure;

[0012] Figure 2 is a flowchart of a signal generation method according to an embodiment of the present disclosure;

[0013] Figure 3 is a flowchart of a signal processing method according to an embodiment of the present disclosure;

[0014] Figure 4 is a structural block diagram of a signal generation apparatus according to an embodiment of the present disclosure;

[0015] Figure 5 is a structural block diagram of a signal processing apparatus according to an embodiment of the present disclosure;

[0016] Figure 6 is a wireless communication system architecture diagram according to an embodiment of the present disclosure;

[0017] Figure 7 is a flowchart of a model training method according to an embodiment of the present disclosure;

[0018] Figure 8 is a framework for training and applying an end-to-end neural network model according to an embodiment of the present disclosure.

[0019] Figure 9 is a second end-to-end neural network model training and application framework according to an embodiment of the present disclosure;

[0020] Figure 10 is a third end-to-end neural network model training and application framework according to an embodiment of the present disclosure. Detailed Implementation

[0021] The embodiments of this disclosure will be described in detail below with reference to the accompanying drawings and examples.

[0022] It should be noted that the terms "first," "second," etc., in the specification, claims, and drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0023] The methods and embodiments provided in this disclosure can be executed in a mobile terminal, a computer terminal, or a similar computing device. Taking a computer terminal as an example, FIG1 is a hardware structure block diagram of a computer terminal in which the methods and embodiments of this disclosure are run. As shown in FIG1, the computer terminal may include one or more (only one is shown in FIG1) processors 102 (processors 102 may include, but are not limited to, microprocessors MCUs or programmable logic devices FPGAs, etc.) and a memory 104 for storing data. The computer terminal may also include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that the structure shown in FIG1 is only illustrative and does not limit the structure of the computer terminal. For example, the computer terminal may also include more or fewer components than shown in FIG1, or have a different configuration than shown in FIG1.

[0024] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the signal generation method in this embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thus implementing the above-described method. The memory 104 may include high-speed random access memory and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to a computer terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0025] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider for the computer terminal. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0026] This embodiment provides a signal generation method running on the aforementioned computer terminal. Figure 2 is a flowchart of the signal generation method according to an embodiment of this disclosure. As shown in Figure 2, the process includes the following steps:

[0027] Step S202: Modulate the data bits to generate data symbols, and modulate the pilot bits to generate pilot symbols;

[0028] In an exemplary embodiment of this disclosure, modulating data bits includes: modulating data bits using an irregular data constellation diagram.

[0029] It should be noted that irregular constellation diagrams are trained and optimized for specific channel conditions, which can improve the anti-interference capability of wireless signals in complex channel environments such as multipath propagation and improve the bit error rate performance of the system under low signal-to-noise ratio conditions.

[0030] In an exemplary embodiment of this disclosure, modulating pilot bits to generate pilot symbols includes: modulating pilot bits according to a standard pilot constellation diagram or an irregular pilot constellation diagram obtained through training to generate pilot symbols.

[0031] In an exemplary embodiment of this disclosure, before modulating the data bits to generate data symbols and modulating the pilot bits to generate pilot symbols, the method further includes: generating an irregular data constellation diagram generation network, an irregular pilot constellation diagram generation network, and a power allocation matrix through end-to-end training.

[0032] In an exemplary embodiment of this disclosure, the generation of an irregular data constellation diagram generation network, an irregular pilot constellation diagram generation network, and a power allocation matrix through end-to-end training includes: generating transmitter data, wherein the transmitter data includes modulated pilot symbols, modulated data symbols, and a power allocation matrix; simulating the transmission process of the transmitter data through a channel based on the transmitter data, obtaining receiver data, and performing energy normalization and integration processing on the receiver data; inputting the energy-normalized and integrated data into an artificial intelligence (AI) receiver, calculating the binary cross-entropy based on the bit values ​​output by the AI ​​receiver and the original bit values ​​to determine the LOSS value; updating the parameters of the irregular data constellation diagram generation network, the irregular pilot constellation diagram generation network, the power allocation matrix, and the AI ​​receiver generation network based on the LOSS value, until the model composed of the irregular data constellation diagram generation network, the power allocation matrix generation network, and the AI ​​receiver converges.

[0033] It should be noted that the purpose of normalization is to improve the numerical stability of the data and provide high-quality input for subsequent network training.

[0034] In an exemplary embodiment of this disclosure, generating modulated pilot symbols includes: selecting a set of fixed-length floating-point complex data according to the pilot modulation order, normalizing the floating-point complex data, generating an irregular pilot constellation diagram; and using the irregular constellation diagram to modulate the pilot bits of the transmitting end to obtain modulated pilot symbols.

[0035] In an exemplary embodiment of this disclosure, generating modulated data symbols includes: selecting a set of fixed-length floating-point complex data according to the data modulation order, normalizing the floating-point complex data, generating an irregular data constellation diagram; and using the irregular constellation diagram to modulate the data bits at the transmitting end to obtain modulated data symbols.

[0036] In an exemplary embodiment of this disclosure, generating a power allocation matrix includes: inputting fixed floating-point data into a power allocation matrix generation network to obtain the corresponding power allocation matrix.

[0037] Step S204: Data symbols and pilot symbols are superimposed on the same time-frequency resources to generate aliased data;

[0038] In an exemplary embodiment of this disclosure, superimposing data symbols and pilot symbols on the same time-frequency resources to generate aliased data includes: determining the power allocation ratio of data symbols and pilot symbols according to a power allocation matrix, and superimposing data symbols and pilot symbols on the same time-frequency resources according to the power allocation ratio to generate aliased data.

[0039] By superimposing data symbols and pilot symbols using a power allocation matrix, the power allocation of data symbols and pilot symbols is dynamically adjusted, avoiding the adverse effects of excessively high or low symbol power on channel estimation and data demodulation.

[0040] Step S204: Obtain the aliasing signal based on the aliasing data and send the aliasing signal to the receiving end.

[0041] In an exemplary embodiment of this disclosure, obtaining an aliased signal based on aliased data includes: performing orthogonal frequency division multiplexing (OFDM) symbol shaping processing on the aliased data and loading a cyclic prefix to obtain the aliased signal to be transmitted; wherein the aliased data is frequency domain data and the aliased signal is a time domain signal.

[0042] It should be noted that the introduction of irregular constellation diagrams provides greater flexibility to this overlay scheme. Irregular constellation diagrams optimize the distribution of constellation points based on channel conditions, allowing the system to utilize channel capacity more efficiently, thereby further improving data transmission efficiency and robustness.

[0043] Figure 3 is a flowchart of a signal processing method according to an embodiment of the present disclosure. As shown in Figure 3, the process includes the following steps:

[0044] Step S302: Receive the aliased signal, convert the aliased signal into a received frequency domain signal, and perform energy normalization processing on the received frequency domain signal to obtain the first signal;

[0045] In an exemplary embodiment of this disclosure, converting an aliased signal into a receiver frequency domain signal includes: simulating the transmission process of the aliased signal through a channel using a frequency domain channel dataset to generate a receiver frequency domain signal.

[0046] Step S304: Integrate the first signal and the pilot signal to obtain the second signal, and input the second signal into the artificial intelligence (AI) receiver to obtain the data symbol bit information and the log-likelihood ratio (LLR) value of the data symbol bit information;

[0047] Step S306: Verify the correctness of AI receiver decoding and evaluate transmission performance based on LLR value.

[0048] In an exemplary embodiment of this disclosure, data integration of a first signal and a pilot signal to obtain a second signal includes: superimposing the real and imaginary parts of the first signal and the real and imaginary parts of the pilot signal on the last dimension of a matrix to obtain the second signal.

[0049] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this disclosure.

[0050] This embodiment also provides a signal generation device and a signal processing device, which are used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0051] Figure 4 is a structural block diagram of a signal generation device according to an embodiment of the present disclosure. As shown in Figure 4, the device includes a first generation module 10, a second generation module 20, and a transmission module 30.

[0052] The first generation module 10 is used to modulate data bits to generate data symbols and to modulate pilot bits to generate pilot symbols.

[0053] The second generation module 20 is used to superimpose data symbols and pilot symbols on the same time-frequency resources to generate aliased data;

[0054] The transmitting module 30 is used to obtain the aliasing signal based on the aliasing data and send the aliasing signal to the receiving end.

[0055] Figure 5 is a structural block diagram of a signal processing apparatus according to an embodiment of the present disclosure. As shown in Figure 5, the apparatus includes a processing module 40, an input module 50, and an evaluation module 60.

[0056] Processing module 40 is used to receive aliased signals, convert aliased signals into received frequency domain signals, and perform energy normalization processing on received frequency domain signals to obtain a first signal.

[0057] The input module 50 is used to integrate the first signal and the pilot signal to obtain the second signal, and input the second signal into the artificial intelligence (AI) receiver to obtain the data symbol bit information and the log-likelihood ratio (LLR) value of the data symbol bit information.

[0058] Evaluation module 60 is used to verify the correctness of AI receiver decoding and evaluate transmission performance based on LLR value.

[0059] It should be noted that the above modules can be implemented by software or hardware. For the latter, they can be implemented in the following ways, but are not limited to: all the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0060] To facilitate understanding of the technical solutions provided in the application embodiments, the following description is based on specific scenario embodiments.

[0061] Figure 6 is a wireless communication system architecture diagram according to an embodiment of the present disclosure. The transmitting end of the system includes three deep learning neural networks: a pilot constellation diagram generation network, a data constellation diagram generation network, and a power allocation matrix generation network.

[0062] The pilot constellation generation network generates an irregular pilot constellation map mapping table through end-to-end training during the training phase. In the application phase, the pilot bits are directly modulated using the irregular constellation map mapping table obtained during the training phase to obtain pilot symbols. The data constellation generation network generates an irregular data constellation map mapping table through end-to-end training during the training phase. In the application phase, the encoded data bits are directly modulated using the irregular constellation map mapping table obtained during the training phase to obtain data symbols. The power allocation matrix generation network generates a power allocation matrix through end-to-end training during the training phase. In the application phase, the data and pilot signals are superimposed using the power allocation matrix obtained during training to obtain a mixed signal on the same time-frequency resources.

[0063] The receiver also employs a deep neural network-based AI receiver. This receiver takes the frequency domain signal as input, and if necessary, the power allocation matrix as input, outputting LLR information of the demodulated data bits. The neural networks of the transmitter and receiver in this system can be trained simultaneously using an end-to-end joint training method, or they can be trained in a step-by-step cascaded manner. This disclosure utilizes the power allocation matrix for data pilot superposition, where the pilots do not independently occupy time-frequency resources, significantly improving spectrum utilization and increasing system throughput. Furthermore, based on the data pilot superposition scheme, an irregular constellation diagram modulation scheme is proposed. The irregular constellation diagram adjusts the constellation point spacing according to channel conditions, enabling the system to flexibly adapt to the channel environment, especially exhibiting superior demodulation performance under low signal-to-noise ratio conditions. In addition, the AI ​​transmitter and receiver obtained through end-to-end training can learn complex channel characteristics and nonlinear effects through deep neural networks, automatically optimizing the signal processing flow. This results in stronger robustness and adaptability in noisy, interference-prone, and even unknown nonlinear distortion environments, achieving efficient signal detection and recovery. In summary, compared with traditional communication systems, this disclosure improves the anti-interference capability and bit error rate performance while ensuring that the system throughput and spectrum utilization efficiency are increased.

[0064] Assuming the system shown in Figure 6 contains an L-stream signal, the transmitting pilot bits X are randomly generated. bit,p Data bit X from the origin bit The training mainly includes channel data collection, transmitter data generation, receiver data acquisition and integration, as well as LOSS calculation and parameter updates.

[0065] Figure 7 is a flowchart of a model training method according to an embodiment of the present disclosure. As shown in Figure 7, the process includes the following steps:

[0066] Step S701, Channel data collection;

[0067] Specifically, before model training, the channel dataset needs to be constructed and organized. This can be done by collecting frequency domain channel data in the field or generating it through simulation, followed by normalization. The purpose of normalization is to improve the numerical stability of the data, providing high-quality input for subsequent network training.

[0068] Step S702, Data generation at the transmitting end;

[0069] Specifically, the transmitting end includes a pilot constellation diagram generation network, a data constellation diagram generation network, and a power allocation matrix generation network. The specific data generation process is as follows:

[0070] Step S7021, pilot signal generation;

[0071] Specifically, the pilot constellation diagram generation network is optional. The pilots can be modulated directly using a standard NR constellation diagram to obtain the modulated pilot symbol X without training. p When using a pilot constellation generation network to generate irregular pilot constellations, assume the pilot modulation order is Q. m,p Select a set of fixed lengths The floating-point complex data is normalized and used as a pilot constellation diagram. Based on this irregular constellation diagram, the transmitting pilot bits X are... bit,p Modulation is performed to obtain the modulated pilot symbol X. p .

[0072] Step S7022, data generation;

[0073] Specifically, assume the data modulation order is Q. m Select a set of fixed lengths The floating-point complex data is normalized and used as a data constellation diagram. Based on this irregular constellation diagram, the originating data bits X are analyzed. bit Modulation is performed to obtain the modulated data symbol X. sym .

[0074] Step S7023, power allocation matrix generation;

[0075] Specifically, a fixed set of floating-point data is selected as the input to the power allocation matrix generation network to obtain the power allocation matrices A1…A1 corresponding to each of the L-stream signals. L The data symbols and pilot symbols are layer-mapped, and then superimposed on the same time-frequency resources according to the power allocation matrix corresponding to the L-stream signal. The superposition process is shown in Formula 1.

[0076] Step S703: Data acquisition and integration at the receiving end;

[0077] Specifically, the frequency domain signal X obtained by superimposing the transmitter data and pilot signals is obtained. mix Then, the frequency domain channel dataset is used to simulate the signal transmission process through the channel, generating the received frequency domain signal Y. mix and for signal Y mix Energy normalization is performed to normalize the maximum value of the signal energy to 1, thus obtaining Y. nor See Formula 2. Y nor =Normal max (Y mix (2)

[0078] Get Y nor Then, data integration is performed, and the integration methods include:

[0079] Integration Method 1: Integrating signal Y nor Real part, imaginary part, pilot signal X p The real and imaginary parts of the matrix are superimposed in the last dimension to obtain the integrated signal Y. in .

[0080] Integration Method 2: Integrating signal Y nor Real part, imaginary part, pilot signal X p The real and imaginary parts of the signal and the power allocation matrix are superimposed on the last dimension of the matrix to obtain the integrated signal Y. in .

[0081] Step S704, LOSS calculation and parameter update;

[0082] Specifically, the transmitter and receiver use an AI neural network, which requires LOSS calculation and parameter updates during the training process.

[0083] LOSS calculation: The signal Y after integration in step S703 in The input data to the AI ​​receiver is processed by the receiver network to obtain the output bit value Y. bit The LOSS value of the model is obtained by calculating the binary cross-entropy based on the bit value output by the AI ​​receiver and the original bit value, as shown in Formula 3. Loss = BCE(X) bit ,Y bit (3)

[0084] Parameter update: Different training architectures have different parameter update methods. The following are three typical training architectures:

[0085] Training Architecture 1: Joint Training (Standard Pilot Constellation Diagram).

[0086] As shown in Figure 8, the pilot signals directly use the standard constellation diagram, and the data constellation diagram generation network and the power allocation matrix generation network are jointly trained.

[0087] Update the parameters of the data constellation generation network, power allocation matrix generation network, and AI receiver network based on the calculated LOSS value. Repeat steps S702 to S704 in the above embodiments until the model converges, and save the obtained irregular data constellation diagram, power allocation matrix indexed by stream number, and AI receiver parameters.

[0088] Training Architecture 2: Joint Training (Irregular Pilot Constellation Graph).

[0089] As shown in Figure 9, the pilot constellation diagram generation network, the data constellation diagram generation network, and the power allocation matrix generation network are jointly trained.

[0090] Update the parameters of the pilot constellation generation network, data constellation generation network, power allocation matrix generation network, and AI receiver network based on the calculated LOSS value. Repeat steps S702 to S704 in the above embodiments until the model converges, and save the obtained irregular pilot constellation diagram, irregular data constellation diagram, power allocation matrix indexed by stream number, and AI receiver parameters.

[0091] Training Architecture 3: Step-by-Step Training.

[0092] As shown in Figure 10, the pilot constellation diagram generation network, the data constellation diagram generation network, and the power allocation matrix generation network are trained step-by-step. It should be noted that the generation order of the networks during step-by-step training is not limited to the following order and can be flexibly adjusted according to actual performance.

[0093] First, fix the data modulation method and power allocation matrix, train the pilot constellation diagram generation network, update the pilot constellation diagram generation network according to the calculated LOSS value, repeat steps S702 to S704 in the above embodiment until the model converges, and save the obtained irregular pilot constellation diagram and AI receiver parameters.

[0094] Then, pilot modulation is performed according to the saved pilot constellation diagram. The data constellation diagram generation network is trained with the newly saved AI receiver parameters as the initial values ​​and the power allocation matrix is ​​fixed. The pilot constellation diagram generation network is updated according to the calculated LOSS value. Steps S702 to S704 are repeated iteratively until the model converges. The obtained irregular data constellation diagram and AI receiver parameters are saved.

[0095] Finally, the model is modulated according to the saved pilot and data constellation diagram. The power allocation matrix generation network is trained using the saved AI receiver parameters as the initial values. The pilot constellation diagram generation network is updated according to the calculated LOSS value. Steps S702 to S704 are repeated until the model converges. The obtained power allocation matrix and AI receiver parameters indexed by stream number are saved.

[0096] This system is based on three generative networks: a pilot constellation diagram (optional), a data constellation diagram, and a power allocation matrix. It also utilizes cascaded or parallel training architectures and, considering the differences in the input format of the receiver model mentioned in step S703, allows for various combinations of training methods, including but not limited to the three typical training architectures described above. Furthermore, the type and structure of the neural network model used at the transceiver end in this embodiment are not limited.

[0097] This embodiment of the disclosure utilizes the pilot constellation diagram, data constellation diagram, and power allocation matrix obtained during the training phase through coordinated processing at the transmitting and receiving ends to complete the generation, transmission, and decoding of the transmitted signal, including the following steps:

[0098] Step S801, Data generation at the transmitting end;

[0099] Specifically, assuming the system contains an L-stream signal, the transmitting pilot bit X... bit,p Modulation is performed according to a standard constellation diagram or a pilot constellation diagram obtained through training to obtain the modulated pilot symbol X. p Originating data bits X bit After channel coding and modulation according to the data constellation diagram obtained through training, the modulated data symbol X is obtained. sym Additionally, based on the power allocation matrix obtained during training, the power allocation matrices A1…A1…A1…A1…A2…A3…A4…A5…A6…A7 ... L The data symbols and pilot symbols are layer-mapped, and then superimposed on the same time-frequency resources according to the power allocation matrix corresponding to the L-stream signal. The superposition process is shown in Equation 1.

[0100] Step S802, signal modulation and transmission;

[0101] Specifically, the superimposed aliased data undergoes OFDM symbol shaping to convert the frequency domain to the time domain, and a cyclic prefix (CP) is added to generate the time-domain signal to be transmitted. The aliasing signal is then sent after the signaling indicates the number of streams.

[0102] Step S803, receiver signal preprocessing;

[0103] Specifically, the receiver removes the cyclic prefix (CP) from the received signal and performs a Fast Fourier Transform (FFT) on the time-domain signal to restore the signal to the frequency domain as input for subsequent processing.

[0104] Step S804: Demodulation and decoding of the receiver signal;

[0105] Specifically, the received frequency domain signal Y is obtained after processing in step S803. mix Based on the number of streams indicated by the received signaling, select the corresponding AI receiver parameters and the power allocation matrix A1…A corresponding to each L stream signal. L and for Y mix Energy normalization is performed to obtain Y nor As shown in Formula 2.

[0106] Then, according to the two integration methods mentioned in the model training, the signal Y... nor Real part, imaginary part, pilot signal X p The real and imaginary parts of the signal and the power allocation matrix (optional) are superimposed on the last dimension of the matrix to obtain the integrated signal Y. in As shown in Formula 3, the integrated signal Y inThe integrated input data of the AI ​​receiver is output as the bit value Y. bit The system integrates the bit decision results to output the final data symbol bit information, and calculates the LLR value to verify the correctness of the receiver decoding and evaluate the transmission performance.

[0107] Scenario Example 1

[0108] Assume an OFDM SISO communication system with a single data stream, 24 scheduled RBs, 288 scheduled REs, and 12 data symbols. The pilot signal uses BPSK modulation, and the modulation order of the data signal is Q. m =4. Channel data is collected through the actual air interface environment. The transceiver process disclosed herein includes model training and model application.

[0109] The model training process includes the following steps:

[0110] Step S901, Channel dataset collection;

[0111] Specifically, a dataset is constructed by collecting frequency domain channel data in the actual air interface environment, and the channel data is normalized.

[0112] Step S901, Data generation at the transmitting end;

[0113] Specifically, this system is a single-stream signal. Randomly generated transmitter pilot bits X bit,p The modulated pilot symbol X is obtained after BPSK modulation. p The dimension is 288*12. Randomly generate the originating data bits X. bit Select a set of fixed lengths The floating-point complex data is normalized and used as a data constellation diagram. Based on this irregular constellation diagram, the originating data bits X are analyzed. bit Modulation is performed to obtain the modulated data symbol X. sym The vectors are 288*12 in dimension. A set of vectors with a random dimension of 288*12 is used as input to the power allocation matrix generation network to obtain the power allocation matrix A1 for a single stream. The pilot symbols and data symbols are layer-mapped and superimposed according to the power allocation matrix A1, as shown in Equation 1, where l = 1.

[0114] Step S903: Data acquisition and integration at the receiving end;

[0115] Specifically, the frequency domain signal X obtained by superimposing the transmitter data and pilot signals is obtained. mix After (dimension 288*12), the frequency domain channel dataset in step S901 is used to simulate the transmission process of the signal through the channel, generating the receiving end frequency domain signal Y. mix The dimension is 288*12. For the signal Y... mixEnergy normalization is performed to normalize the maximum value of the signal energy to 1, thus obtaining Y. nor See Formula 2.

[0116] Signal Y nor Real part, imaginary part, pilot signal X p The real and imaginary parts of the matrix are superimposed in the last dimension to obtain the integrated signal Y. in The dimension is 288*12*4 (where 4 includes the real and imaginary parts of the normalized signal and pilot).

[0117] Step S904: Loss calculation and parameter update;

[0118] Specifically, the integrated signal Y from step S903 in The input data to the single-stream AI receiver is processed by the receiver network to obtain the output bit value Y. bit The LOSS value of the model is obtained by calculating the binary cross-entropy based on the bit value output by the AI ​​receiver and the original bit value, as shown in Formula 3.

[0119] Update the parameters of the data constellation generation network, power allocation matrix generation network, and AI receiver network based on the calculated LOSS value. Repeat steps S902 to S904 until the model converges, and save the obtained irregular data constellation diagram, power allocation matrix A1, and single-stream AI receiver parameters.

[0120] The model application part includes the following steps:

[0121] Step S1001, Data generation at the transmitting end;

[0122] Specifically, this system is a single-stream signal. Transmitter pilot bit X bit,p Modulation was performed according to BPSK to obtain the modulated pilot symbol X. p The dimension is 288*12. The originating data bits are X. bit After channel coding and according to the Q obtained during training m =4 data constellation diagram is modulated to obtain the modulated data symbol X sym The dimension is 288*12. Additionally, a single-stream power allocation matrix A1, with a dimension of 288*12, has been obtained during training. Data symbols and pilot symbols are layer-mapped, and based on the power allocation matrix A1, data symbols and pilot symbols are superimposed on the same time-frequency resources. The superposition process is shown in Equation 1.

[0123] Step S1002, signal modulation and transmission;

[0124] Specifically, the superimposed aliased data undergoes OFDM symbol shaping to convert the frequency domain to the time domain, and a cyclic prefix (CP) is added to generate the time-domain signal to be transmitted. The aliasing signal is then sent after the signaling indicates the number of streams.

[0125] Step S1003, preprocessing of the receiver signal;

[0126] Specifically, the receiver removes the cyclic prefix (CP) from the received signal and performs a Fast Fourier Transform (FFT) on the time-domain signal to restore the signal to the frequency domain as input for subsequent processing.

[0127] Step S1004: Demodulation and decoding of the receiver signal;

[0128] Specifically, the receiver frequency domain signal Y with a dimension of 288*12 is obtained after processing in step S1003. mix The AI ​​receiver parameters for training have been acquired from a single stream, and are applied to Y. mix Energy normalization is performed to obtain Y nor As shown in Formula 2.

[0129] Signal Y nor Real part, imaginary part, pilot signal X p The real and imaginary parts of the matrix are superimposed in the last dimension to obtain the integrated signal Y. in The dimension is 288*12*4 (where 4 includes the real and imaginary parts of the normalized signal and pilot). As shown in Equation 3, the integrated signal Y in The input data to the AI ​​receiver is used to obtain the output bit value Y. bit The system integrates the bit decision results to output the final data symbol bit information, and calculates the LLR value to verify the correctness of the receiver decoding and evaluate the transmission performance.

[0130] Scenario Example 2

[0131] Assume an OFDM MIMO communication system with two data streams, 12 scheduled RBs, 144 scheduled REs, and 12 data symbols. The pilot signal modulation order Q... m,p =1, the modulation order Q of the data signal m =6. Channel data is collected through a simulation environment. The transceiver process disclosed herein includes model training and model application.

[0132] The model training process includes the following steps:

[0133] Step S1101: Collect and organize the channel dataset;

[0134] Specifically, a dataset is constructed by collecting frequency domain channel data in a simulation environment, and the channel data is normalized.

[0135] Step S1102, Data generation at the transmitting end;

[0136] Specifically, this system is a two-stream signal. Randomly generated transmitting pilot bits X bit,p Select a set of fixed lengths The floating-point complex data, after normalization, is used as a pilot constellation diagram. Based on this irregular constellation diagram, the transmitting pilot bits X are... bit,p Modulation is performed to obtain the modulated pilot symbol X. p The dimension is 144*12*2. Randomly generate the originating data bits X. bit Select a set of fixed lengths The floating-point complex data, after normalization, is used as a data constellation diagram. Based on this irregular constellation diagram, the originating data bits X are analyzed. bit Modulation is performed to obtain the modulated data symbol X. sym The dimensions are 144*12*2. Two sets of vectors with dimensions of 144*12 are randomly generated as inputs to the power allocation matrix generation network to obtain the power allocation matrices A1 and A2 for the two streams. The pilot symbols and data symbols are layer-mapped and superimposed according to the power allocation matrices A1 and A2, as shown in Equation 1, where l = 1, 2.

[0137] Step S1103: Data acquisition and integration at the receiving end;

[0138] Specifically, the frequency domain signal X obtained by superimposing the transmitter data and pilot signals is obtained. mix After (dimensions of 144*12*2), the frequency domain channel dataset in step S1101 is used to simulate the transmission process of the signal through the channel, generating the received frequency domain signal Y. mix The dimensions are 144*12*2. For the signal Y... mix Energy normalization is performed to normalize the maximum value of the signal energy to 1, thus obtaining Y. nor See Formula 2.

[0139] Signal Y nor Real part, imaginary part, pilot signal X p The real and imaginary parts of the signal and the power allocation matrix are superimposed on the last dimension of the matrix to obtain the integrated signal Y. in The dimensions are 144*12*2*5 (where 5 includes the real and imaginary parts of the normalized signal and pilot, as well as the power factor).

[0140] Step S1104, LOSS calculation and parameter update;

[0141] Specifically, the integrated signal Y from step S1103 is used. in The input data to the two-stream AI receiver is processed by the receiver network to obtain the output bit value Y.bit The LOSS value of the model is obtained by calculating the binary cross-entropy based on the bit value output by the AI ​​receiver and the original bit value, as shown in Formula 3.

[0142] Update the parameters of the pilot constellation generation network, data constellation generation network, power allocation matrix generation network, and AI receiver network based on the calculated LOSS value. Repeat steps S1102 to S1104 until the model converges, and save the obtained irregular pilot constellation diagram, irregular data constellation diagram, power allocation matrices A1 and A2, and two-stream AI receiver parameters.

[0143] Model application includes the following steps:

[0144] Step S1201: Data generation at the transmitting end;

[0145] Specifically, this system is a two-stream signal. Transmitter pilot bit X bit,p According to the Q obtained during training m,p =1 pilot constellation diagram is modulated to obtain modulated pilot symbol X p The dimension is 144*12*2. The originating data bits are X. bit After channel coding and according to the Q obtained during training m =6 data constellation diagram is modulated to obtain the modulated data symbol X sym The dimensions are 144*12*2. Additionally, power allocation matrices A1 and A2 for the two streams have been obtained during training, each with dimensions of 144*12. Data symbols and pilot symbols are layer-mapped, and based on power allocation matrices A1 and A2, data symbols and pilot symbols are superimposed on the same time-frequency resources, as shown in Equation 1.

[0146] Step S1202, signal modulation and transmission;

[0147] Specifically, the superimposed aliased data undergoes OFDM symbol shaping to convert the frequency domain to the time domain, and a cyclic prefix (CP) is added to generate the time-domain signal to be transmitted. The aliasing signal is then sent after the signaling indicates the number of streams.

[0148] Step S1203, preprocessing of the receiver signal;

[0149] Specifically, the receiver removes the cyclic prefix (CP) from the received signal and performs a Fast Fourier Transform (FFT) on the time-domain signal to restore the signal to the frequency domain as input for subsequent processing.

[0150] Step S1204: Demodulation and decoding of the receiver signal;

[0151] Specifically, the receiving frequency domain signal Y with dimensions of 144*12*2 is obtained after processing in step S1203.mix The training has acquired the parameters of the two-stream AI receiver and the two-stream power allocation matrices A1 and A2, for Y. mix Energy normalization is performed to obtain Y nor As shown in Formula 2.

[0152] Signal Y nor Real part, imaginary part, pilot signal X p The real and imaginary parts of the signal and the power allocation matrix are superimposed on the last dimension of the matrix to obtain the integrated signal Y. in The dimension is 144*12*2*5 (where 5 includes the real and imaginary parts of the normalized signal and pilot, as well as the power factor). As shown in Equation 3, the integrated signal Y... in The input data to the AI ​​receiver is used to obtain the output bit value Y. bit The system integrates the bit decision results to output the final data symbol bit information, and calculates the LLR value to verify the correctness of the receiver decoding and evaluate the transmission performance.

[0153] Scenario Example 3

[0154] Assume an OFDM MIMO communication system with four data streams, 6 scheduled RBs, 72 scheduled REs, and 12 data symbols. The pilot signal modulation order Q... m,p =2, the modulation order Q of the data signal m =8, channel data is collected through a simulation environment, and the transceiver process disclosed herein includes model training and model application.

[0155] Model training includes the following steps:

[0156] Step S1301: Collect and organize the channel dataset;

[0157] Specifically, a dataset is constructed by collecting frequency domain channel data in a simulation environment, and the channel data is normalized.

[0158] Step S1302, Data generation at the transmitting end;

[0159] Specifically, this system is a four-stream signal. The transmitting pilot bits X are randomly generated. bit,p Select a set of fixed lengths The floating-point complex data, after normalization, is used as a pilot constellation diagram. Based on this irregular constellation diagram, the transmitting pilot bits X are... bit,p Modulation is performed to obtain the modulated pilot symbol X. p The dimensions are 72*12*4. Randomly generate the originating data bits X. bit Select a set of fixed lengths The floating-point complex data, after normalization, is used as a data constellation diagram. Based on this irregular constellation diagram, the originating data bits X are analyzed. bit Modulation is performed to obtain the modulated data symbol X. sym The dimensions are 72*12*4. Four sets of 72*12 vectors are randomly generated as inputs to the power allocation matrix generation network, resulting in power allocation matrices A1, A2, A3, and A4 for the four streams. Pilot symbols and data symbols are layer-mapped and superimposed according to the power allocation matrices A1, A2, A3, and A4, as shown in Equation 1, where l = 1, 2, 3, 4.

[0160] Step S1303: Data acquisition and integration at the receiving end;

[0161] Specifically, the frequency domain signal X obtained by superimposing the transmitter data and pilot signals is obtained. mix After (dimensions 72*12*4), the frequency domain channel dataset in step S1301 is used to simulate the transmission process of the signal through the channel, generating the received frequency domain signal Y. mix The dimensions are 72*12*4. For the signal Y... mix Energy normalization is performed to normalize the maximum value of the signal energy to 1, thus obtaining Y. nor As shown in Formula 2.

[0162] Signal Y nor Real part, imaginary part, pilot signal X p The real and imaginary parts of the signal and the power allocation matrix are superimposed on the last dimension of the matrix to obtain the integrated signal Y. in The dimensions are 72*12*4*5 (where 5 includes the real and imaginary parts of the normalized signal and pilot, as well as the power factor).

[0163] Step S1304: Loss calculation and parameter update;

[0164] Specifically, the integrated signal Y from step S1303 is used. in The input data to the fourth-stream AI receiver is processed by the receiver network to obtain the output bit value Y. bit The LOSS value of the model is obtained by calculating the binary cross-entropy based on the bit value output by the AI ​​receiver and the original bit value, as shown in Formula 3.

[0165] First, fix the data modulation method and power allocation matrix, train the pilot constellation diagram generation network, update the pilot constellation diagram generation network according to the LOSS value, repeat the iteration from step S1302 to step S1304 until the model converges, and save the obtained irregular pilot constellation diagram and four-stream AI receiver parameters.

[0166] Then, pilot modulation is performed according to the saved pilot constellation diagram. The data constellation diagram generation network is trained with the newly saved AI receiver parameters as the initial values ​​and the power allocation matrix is ​​fixed. The pilot constellation diagram generation network is updated according to the LOSS value. Steps S1302 to S1304 are repeated until the model converges. The obtained irregular data constellation diagram and four-stream AI receiver parameters are saved.

[0167] Finally, the modulation is performed according to the saved pilot and data constellation diagrams. The power allocation matrix generation network is trained with the newly saved AI receiver parameters as the initial values. The pilot constellation diagram generation network is updated according to the LOSS value. Steps S1302 to S1304 are repeated until the model converges. The obtained power allocation matrices A1, A2, A3, A4 and the four-stream AI receiver parameters are saved.

[0168] Model application includes the following steps:

[0169] Step S1401, Data generation at the transmitting end;

[0170] Specifically, this system is a four-stream signal. Transmitting pilot bit X bit,p According to the Q obtained during training m,p =2 pilot constellation diagram is modulated to obtain the modulated pilot symbol X. p The dimensions are 72*12*4. The originating data bits are X. bit After channel coding and according to the Q obtained during training m =8 data constellation diagram is modulated to obtain the modulated data symbol X sym The dimensions are 72*12*4. Additionally, the power allocation matrices A1, A2, A3, and A4 for the four streams were obtained during training, each with a dimension of 72*12. Data symbols and pilot symbols are layer-mapped, and based on the power allocation matrices A1, A2, A3, and A4, data symbols and pilot symbols are superimposed on the same time-frequency resources, as shown in Equation 1.

[0171] Step S1402, signal modulation and transmission;

[0172] Specifically, the superimposed aliased data undergoes OFDM symbol shaping to convert the frequency domain to the time domain, and a cyclic prefix (CP) is added to generate the time-domain signal to be transmitted. The aliasing signal is then sent after the signaling indicates the number of streams.

[0173] Step S1403, preprocessing of the receiver signal;

[0174] Specifically, the receiver removes the cyclic prefix (CP) from the received signal and performs a Fast Fourier Transform (FFT) on the time-domain signal to restore the signal to the frequency domain as input for subsequent processing.

[0175] Step S1403: Demodulation and decoding of the receiver signal;

[0176] Specifically, the receiving frequency domain signal Y with dimensions of 72*12*4 is obtained after processing in step S1403. mix The training has acquired the parameters of the four-stream AI receiver and the four-stream power allocation matrices A1, A2, A3, and A4, for Y. mix Energy normalization is performed to obtain Y nor As shown in Formula 2.

[0177] Signal Y nor Real part, imaginary part, pilot signal X p The real and imaginary parts of the signal and the power allocation matrix are superimposed on the last dimension of the matrix to obtain the integrated signal Y. in The dimensions are 72*12*4*5 (where 5 includes the real and imaginary parts of the normalized signal and pilot, as well as the power factor). As shown in Equation 3, the integrated signal Y... in The input data to the AI ​​receiver is used to obtain the output bit value Y. bit The system integrates the bit decision results to output the final data symbol bit information, and calculates the LLR value to verify the correctness of the receiver decoding and evaluate the transmission performance.

[0178] Compared to traditional methods, the data pilot overlay scheme superimposes pilot signals onto data signals according to a power allocation matrix, transmitting both pilot and data signals simultaneously on the same time-frequency resources. This avoids the inefficiency caused by the isolation of pilot and data signals in time-frequency resources in traditional schemes. Then, an AI receiver trained end-to-end separates the pilot signals from the superimposed signals and performs channel characteristic estimation. Compared to traditional methods, data pilot overlay is significantly more advantageous in scenarios with limited spectrum resources or high requirements for spectrum utilization.

[0179] Furthermore, the introduction of irregular constellation diagrams provides greater flexibility to this superposition scheme. Irregular constellation diagrams optimize the distribution of constellation points according to channel conditions, allowing the system to utilize channel capacity more efficiently, thereby further improving data transmission efficiency and robustness. Traditional regular constellation diagrams, such as QPSK or 16-QAM, have uniform constellation point distributions. While simple to implement, they may not maximize transmission performance under poor channel conditions. By combining irregular constellation diagrams with data pilot superposition, the system can achieve higher anti-interference capabilities and better bit error rate performance, thus providing a superior technical solution for future high-density, high-speed wireless communication systems.

[0180] In summary, the embodiments of this disclosure, by introducing an irregular constellation diagram based on data pilot superposition, not only retain efficient spectrum utilization but also further improve the system's transmission performance in complex channel environments, solving problems such as low resource utilization and insufficient anti-interference capability in traditional communication systems. This innovative technology is of great significance for future new communication systems, providing better support for reliable communication in high-speed mobile scenarios.

[0181] Embodiments of this disclosure also provide a computer-readable storage medium storing a computer program configured to perform the steps in any of the above method embodiments when executed.

[0182] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0183] Embodiments of this disclosure also provide an electronic device including a memory and a processor, the memory storing a computer program and the processor being configured to run the computer program to perform the steps in any of the above method embodiments.

[0184] In one exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0185] Specific examples in this embodiment can be found in the examples described in the above embodiments and exemplary implementations, and will not be repeated here.

[0186] It is obvious to those skilled in the art that the modules or steps of this disclosure described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. They can be implemented using computer-executable program code, and thus can be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, this disclosure is not limited to any particular combination of hardware and software.

[0187] The above description is merely a preferred embodiment of this disclosure and is not intended to limit this disclosure. Various modifications and variations can be made to this disclosure by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A signal generation method, applied at a transmitting end, comprising: Modulate the data bits to generate data symbols, and modulate the pilot bits to generate pilot symbols; The data symbols and the pilot symbols are superimposed on the same time-frequency resources to generate aliased data; The aliasing signal is obtained based on the aliasing data, and the aliasing signal is sent to the receiving end.

2. The method according to claim 1, wherein, The modulation of data bits includes: modulating data bits using an irregular data constellation diagram.

3. The method according to claim 1, wherein, The step of superimposing the data symbols and the pilot symbols on the same time-frequency resources to generate aliased data includes: The power allocation ratio of the data symbols and the pilot symbols is determined according to the power allocation matrix. The data symbols and the pilot symbols are superimposed on the same time-frequency resources according to the power allocation ratio to generate aliased data.

4. The method according to claim 1, wherein, The modulation of the pilot bits to generate pilot symbols includes: The pilot bits are modulated according to a standard pilot constellation diagram or an irregular pilot constellation diagram obtained through training to generate the pilot symbols.

5. The method according to claim 1, wherein, The step of obtaining the aliasing signal based on the aliasing data includes: The aliased data is subjected to orthogonal frequency division multiplexing (OFDM) symbol shaping processing, and a cyclic prefix is ​​loaded to obtain the aliased signal to be transmitted. The aliased data is frequency domain data, and the aliased signal is time domain signal.

6. The method according to claim 2, wherein, Before modulating the data bits to generate data symbols, modulating the pilot bits to generate pilot symbols, the process further includes: generating the irregular data constellation diagram generation network, the irregular pilot constellation diagram generation network, and the power allocation matrix through end-to-end training.

7. The method according to claim 6, wherein, The method of generating irregular data constellation diagrams, irregular pilot constellation diagrams, and power allocation matrices through end-to-end training includes: Generate transmitter data, wherein the transmitter data includes modulated pilot symbols, modulated data symbols, and a power allocation matrix; Based on the transmitter data, simulate the transmission process of the transmitter data through the channel, obtain the receiver data, and perform energy normalization and integration processing on the receiver data; The data, after energy normalization and integration processing, is input into an artificial intelligence (AI) receiver. The binary cross-entropy is calculated based on the bit value output by the AI ​​receiver and the original bit value to determine the LOSS value. The parameters of the irregular data constellation generation network, the irregular pilot constellation generation network, the power allocation matrix, and the AI ​​receiver generation network are updated based on the LOSS value until the model composed of the irregular data constellation generation network, the power allocation matrix generation network, and the AI ​​receiver converges.

8. The method according to claim 7, wherein, The generation of modulated pilot symbols includes: Select a set of fixed-length floating-point complex data according to the pilot modulation order, normalize the floating-point complex data, and generate an irregular pilot constellation diagram. The irregular constellation diagram is used to modulate the pilot bits at the transmitting end to obtain the modulated pilot symbols.

9. The method according to claim 7, wherein, The generation of modulated data symbols includes: Select a set of fixed-length floating-point complex data based on the data modulation order, normalize the floating-point complex data, and generate an irregular data constellation diagram; The irregular constellation diagram is used to modulate the data bits at the transmitting end to obtain modulated data symbols.

10. The method according to claim 7, wherein, The process of generating the power allocation matrix includes: inputting fixed floating-point data into the power allocation matrix generation network to obtain the corresponding power allocation matrix.

11. A signal processing method, applied at a receiving end, comprising: The system receives an aliased signal, converts the aliased signal into a receiving frequency domain signal, and performs energy normalization processing on the receiving frequency domain signal to obtain a first signal. The first signal and the pilot signal are integrated to obtain the second signal. The second signal is then input into an artificial intelligence (AI) receiver to obtain the data symbol bit information and the log-likelihood ratio (LLR) value of the data symbol bit information. The correctness of the AI ​​receiver's decoding is verified and the transmission performance is evaluated based on the LLR value.

12. The method according to claim 11, wherein, The step of converting the aliased signal into a received frequency domain signal includes: The frequency domain channel dataset is used to simulate the transmission process of the aliased signal through the channel, thereby generating the frequency domain signal at the receiving end.

13. The method according to claim 11, wherein, The step of integrating the first signal and the pilot signal to obtain the second signal includes: The real and imaginary parts of the first signal and the real and imaginary parts of the pilot signal are superimposed on the last dimension of the matrix to obtain the second signal.

14. A computer-readable storage medium, wherein, The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, it implements the steps of the method described in any one of claims 1 to 10, or implements the steps of the method described in any one of claims 11 to 13.

15. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of the method of any one of claims 1 to 10, or implements the steps of the method of any one of claims 11 to 13.

16. A computer program product comprising a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1 to 10, or implements the steps of the method according to any one of claims 11 to 13.

Images