CN117040683A - Multi-user semantic information transmission method and system based on non-orthogonal multiple access - Google Patents

Abstract

The application provides a multi-user semantic information transmission method and system based on non-orthogonal multiple access, comprising the following steps: acquiring first user information and second user information, encoding the first user information and the second user information, and determining a first characteristic symbol and a second characteristic symbol; the first user information is central user information, and the second user information is edge user information; determining a first distance value between a first user and a base station, a second distance value between a second user and the base station, and determining a first channel gain and a second channel gain according to the first distance value and the second distance value; according to the first characteristic symbol, the second characteristic symbol, the first channel gain and the second channel gain, carrying out non-orthogonal superposition on the first user information and the second user information, and determining a superposition signal; and decoding according to the superposition signal, the first channel gain and the second channel gain, and determining first reconstruction information and second reconstruction information.

Description

Multi-user semantic information transmission method and system based on non-orthogonal multiple access

Technical field

The present application relates to the field of semantic communication technology, and in particular to a multi-user semantic information transmission method and system based on non-orthogonal multiple access access.

Background technique

NOMA (Non-Orthogonal Multiple Access) technology realizes resource reuse by allocating different powers to multiple users on the same time-frequency resource to achieve the purpose of non-orthogonal multiple access. Semantic transmission is a new architecture that integrates user information meaning and information-related features into the communication process. Different from traditional communication, semantic communication does not focus on the accuracy of transmission of individual communication symbols, but aims at conveying meaning and conducting efficient and intelligent communication between users. In the existing technology, communication is performed through a single-user wireless transmission system combining traditional syntax with NOMA technology or combining an autoencoder model and OFDM technology.

However, there is an information interference problem in the existing technology of non-orthogonal transmission, which causes strong interference in information reconstruction at the receiving end, causing the reconstructed information to be non-referential.

Contents of the invention

In view of this, the purpose of this application is to propose a multi-user semantic information transmission method and system based on non-orthogonal multiple access access.

Based on the above purpose, this application provides a multi-user semantic information transmission method based on non-orthogonal multiple access, including: a coding module and a decoding module;

Obtain the first user information and the second user information, encode the first user information and the second user information, and determine the first characteristic symbol and the second characteristic symbol; wherein the first user information is the central user Information, the second user information is edge user information;

Determine a first distance value between the first user and the base station, a second distance value between the second user and the base station, and determine the first channel gain and the second channel gain based on the first distance value and the second distance value;

According to the first characteristic symbol, the second characteristic symbol, the first channel gain and the second channel gain, perform non-orthogonal superposition on the first user information and the second user information to determine Overlay signals;

First reconstruction information and second reconstruction information are determined according to the superimposed signal, the first channel gain and the second channel gain.

Optionally, encoding the first user information and the second user information to determine the first characteristic symbol and the second characteristic symbol includes:

Encode the first user information and the second user information to determine a first time domain signal and a second time domain signal;

Feature extraction is performed on the first user information and the second user information according to the first time domain signal and the second time domain signal, and the first characteristic symbol and the second characteristic symbol are determined.

Optionally, the method also includes:

Perform power normalization on the first characteristic symbol and the second characteristic symbol to determine the first transmission signal of the first user and the second transmission signal of the second user; wherein the first transmission signal is the first transmission signal. The user transmits a signal on a specific carrier, and the second transmission signal is a signal transmitted by the second user on a specific carrier;

The superposition signal is determined based on the first transmission signal, the second transmission signal, the first channel gain and the second channel gain.

Optionally, the power normalization includes:

Obtain the user priority, the first distance value and the second distance value, and determine the location of the first user or the second user on a specific carrier based on the user priority and the first distance value or the second distance value. Send a signal.

Optionally, the first distance value is the distance between the first user and the base station, and the second distance value is the distance between the second user and the base station.

Optionally, the first channel gain is the channel gain between the first user and the base station, and the second channel gain is the channel gain between the second user and the base station;

Determining the first channel gain and the second channel gain according to the first distance value and the second distance value includes:

Determine the fading path loss of the first user according to the first distance value;

Determine the fading path loss of the second user according to the second distance value;

According to the fading path loss of the first user and the fading path loss of the second user, the first channel gain and the second channel gain are determined by the following formula:

Among them, h _m,k is the small-scale fading of user m on carrier k, ρ _m is the fading path loss of user m, m=1,2,3,...,M.

Optionally, determining the first reconstruction information and the second reconstruction information according to the superimposed signal, the first channel gain and the second channel gain includes:

According to the superimposed signal and the first channel gain, perform semantic feature recovery and information reconstruction through a convolutional neural network, and determine the first reconstruction information;

Perform signal modulation and power control on the first reconstructed information to determine the receiving end signal corresponding to the first user signal;

According to the receiving end signal, the superimposed signal and the second channel gain, semantic feature recovery and information reconstruction are performed through a convolutional neural network to determine the second reconstructed information.

Optionally, the method also includes:

Perform channel equalization processing on the superimposed signal through minimum mean square error, perform semantic feature recovery and information reconstruction through a convolutional neural network based on the first channel gain and the superimposed signal after channel equalization processing, and determine the first 1. Reconstruction information.

Optionally, performing semantic feature recovery and information reconstruction through a convolutional neural network according to the receiving end signal, the superimposed signal and the second channel gain, and determining the second reconstruction information includes:

Perform interference elimination on the receiving end signal according to the superimposed signal to determine that there is no interference signal;

Perform channel equalization processing on the interference-free signal through minimum mean square error, and perform decoding according to the second channel gain and the interference-free signal after channel equalization processing to determine the second reconstruction information.

Based on the same inventive concept, embodiments of the present application also provide a multi-user semantic information transmission system based on non-orthogonal multiple access, including:

Coding module, transmitter, multi-carrier channel, receiver and interference cancellation module;

The transmitter includes the encoding module, which includes a first encoder, a second encoder, and a power control module. The transmitter encodes the first user information and the second user information through the encoding module. , determine the first characteristic symbol and the second characteristic symbol corresponding to the first user information and the second user information, perform power normalization on the first characteristic symbol and the second characteristic symbol, and The normalized first characteristic symbol and the second characteristic symbol are sent to the multi-carrier channel as transmission signals;

The multi-carrier channel is used to perform non-orthogonal transmission of the first characteristic symbol and the second characteristic symbol, and determine a first distance value between the first user and the base station, and a second distance value between the second user and the base station, According to the first distance value and the second distance value, a first channel gain and a second channel gain are determined, and the first channel gain and the second channel gain are sent to the receiver and the Interference cancellation module;

The receiver includes the interference cancellation module, which includes a first decoder, a second decoder, and a third encoder. The receiver is configured to receive the first channel gain and the The second channel gain is based on the first characteristic symbol and the second characteristic symbol, the first channel gain and the second channel gain, and the first user information and the second user information. Superimpose, determine the superimposed signal, perform interference elimination and information reconstruction on the superimposed signal, and determine the reconstructed information.

As can be seen from the above, this application provides a multi-user semantic information transmission method and system based on non-orthogonal multiple access. The multi-user semantic information transmission method based on non-orthogonal multiple access includes: Obtain the first user information and the second user information, encode the first user information and the second user information, and determine the first characteristic symbol and the second characteristic symbol; wherein the first user information is the central user Information, the second user information is edge user information; determine the first distance value between the first user and the base station, and the second distance value between the second user and the base station, according to the first distance value and the second distance value , determine the first channel gain and the second channel gain; according to the first characteristic symbol, the second characteristic symbol, the first channel gain and the second channel gain, the first user information and the The second user information is non-orthogonally superimposed to determine a superimposed signal; decoding is performed according to the superimposed signal, the first channel gain and the second channel gain to determine the first reconstruction information and the second reconstruction information. This application designs different processing procedures for central users and edge users respectively to reduce interference and separate the users' respective semantic information for reconstruction. This proposal improves the information carrying capacity under a specific bandwidth through data dimensionality reduction and compression, also increases the number of user accesses, and uses the correlation characteristics of multiple users to improve the quality of information reconstruction.

Description of the drawings

In order to more clearly illustrate the technical solutions in this application or related technologies, the drawings needed to be used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings in the following description are only for the purposes of this application. Embodiments, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without exerting creative efforts.

Figure 1 is a schematic flow chart of a multi-user semantic information transmission method based on non-orthogonal multiple access according to an embodiment of the present application;

Figure 2 is a schematic diagram of a multi-user semantic information transmission method based on non-orthogonal multiple access according to an embodiment of the present application;

Figure 3 is a schematic diagram of the encoding module of the multi-user semantic information transmission method based on non-orthogonal multiple access according to the embodiment of the present application;

Figure 4 is a schematic diagram of a transmission module of an embodiment of a multi-user semantic information transmission method based on non-orthogonal multiple access according to the embodiment of the present application;

Figure 5 is a schematic diagram of the interference cancellation module of the multi-user semantic information transmission method based on non-orthogonal multiple access according to the embodiment of the present application;

Figure 6 is a schematic diagram of a multi-user semantic information transmission system based on non-orthogonal multiple access according to an embodiment of the present application;

Figure 7 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clear, the present application will be further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings.

It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of this application should have the usual meanings understood by those with ordinary skills in the field to which this application belongs. The "first", "second" and similar words used in the embodiments of this application do not indicate any order, quantity or importance, but are only used to distinguish different components. Words such as "include" or "comprising" mean that the elements or things appearing before the word include the elements or things listed after the word and their equivalents, without excluding other elements or things. Words such as "connected" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right", etc. are only used to express relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

As mentioned in the background technology section, NOMA (Non-Orthogonal Multiple Access) technology implements resource multiplexing by allocating different powers to multiple users on the same time-frequency resource to achieve non-orthogonal multiple access. The purpose of multiple access. Semantic transmission is a new architecture that integrates user information meaning and information-related features into the communication process. Different from traditional communication, semantic communication does not focus on the accuracy of transmission of individual communication symbols, but aims at conveying meaning and conducting efficient and intelligent communication between users. In the existing technology, the traditional syntax is combined with NOMA (Non-Orthogonal Multiple Access, non-orthogonal multiple access technology) technology or by combining the autoencoder model and OFDM (Orthogonal Frequency Division Multiplexing, orthogonal frequency division multiplexing technology) technology. Single user wireless transmission system for communication. Among them, NOMA technology has significant advantages in increasing user access and improving spectrum efficiency by enhancing spectrum reuse. The autoencoder is an unsupervised learning model that can use input information as a learning target and perform representation learning on the meaning of the input information. The autoencoder is an important tool that combines NOMA with user semantic communication and plays a role in semantic communication systems. It plays a key role in extracting and reconstructing user semantic information.

In the existing technology, the traditional syntax communication method combining NOMA technology, the semantic communication method using OFDM technology and user semantic communication using the autoencoder model, and the semantic communication method combining NOMA technology and the autoencoder model for discrete modulation and demodulation are generally adopted. However, the traditional syntax communication method combined with NOMA technology uses traditional QAM modulation, there is no correlation between modulated signals, and the user's syntax information does not have low-dimensional features, requiring a large amount of information to be transmitted. Using the autoencoder model to combine OFDM technology with user semantic communication methods and using orthogonal multiple access technology cannot maximize the utilization of spectrum resources and ignore the correlation characteristics of information between users. Existing research on semantic communication methods that combines NOMA technology and autoencoder models has designed a quantizer to discretize semantic information. Some semantic features will be lost during the discretization process. The transmission of continuous-valued multi-user semantic information has not yet been carried out. design.

In view of this, embodiments of the present application provide a multi-user semantic information transmission method and system based on non-orthogonal multiple access access. The multi-user semantic information transmission method uses the deep learning autoencoder as a tool to design a communication system that combines NOMA technology with multi-user semantic information transmission. The overall architecture of the system is to equip a set of codecs at the transmitting end and receiving end respectively. Each pair of encoders is equipped on the central user 1 and edge user 2 based on NOMA transmission, and the decoder is equipped on the corresponding receiving end (base station). By learning the semantic information and correlation features of each pair of users, the system can enhance the quality of information reconstruction at the receiving end and increase user access. The advantage of this invention is that it realizes multi-user semantic access based on NOMA technology, uses the correlation of semantic information between users to extract effective semantic features, and can effectively improve user access compared with the Orthogonal Multiple Access (OMA) scheme. Input quantity and information reconstruction quality.

As shown in Figure 1, the multi-user semantic information transmission method includes:

Step S102: Obtain the first user information and the second user information, encode the first user information and the second user information, and determine the first characteristic symbol and the second characteristic symbol; wherein, the first user information is central user information, and the second user information is edge user information;

Step S104: Determine the first distance value between the first user and the base station, and the second distance value between the second user and the base station. Based on the first distance value and the second distance value, determine the first channel gain and the second channel gain. gain;

Step S106: Perform non-orthogonal processing on the first user information and the second user information according to the first characteristic symbol, the second characteristic symbol, the first channel gain and the second channel gain. Superposition, determine the superposition signal;

Step S108: Determine first reconstruction information and second reconstruction information according to the superimposed signal, the first channel gain and the second channel gain.

As shown in Figure 2, it is a structural diagram of the non-orthogonal multiple access multi-user semantic information transmission system of the present application. According to Figure 2, it can be seen that this system considers a communication process in which multiple users send information to the base station. The main process of the system is: the transmitter, that is, the transmitter first uses a pair of NOMA encoders to process the information sent by the central user 1 and the edge user 2. Feature extraction is performed to obtain feature symbols, in which the first feature symbol is determined by feature extraction on the first user information sent by center user 1, and the second feature symbol is determined by feature extraction on the second user information sent by edge user 2. . It should be noted that the center users are users who are close to the base station, and the edge users are users who are far from the base station.

It can be understood that the "1, 2" in the above "center user 1" and "edge user 2" are only used to distinguish center users and edge users and have no actual technical significance. The center user can also be marked as 2. , edge users can also be marked as 1.

As an implementation manner, the central user is 1 and the edge user is 2. When there are multiple edge users, there may be edge user 3, edge user 4, ..., and edge user n.

Further, after the first characteristic symbol and the second characteristic symbol are determined, power normalization control is performed on the first characteristic symbol and the second characteristic symbol, and then the first characteristic symbol and the second characteristic symbol after the power normalization control are The symbols are sent to multi-carrier signals for non-orthogonal superposition, and the superimposed signals are determined and sent to the receiver. The receiver receives the superimposed signals of central user 1 and edge user 2, and uses the designed semantic interference elimination mechanism to detect user 1 and user 1 respectively. The message sent by User 2 is reconstructed. The following describes in detail the three processes of designing a pair of encoders for central user 1 and edge user 2 at the transmitting end, non-orthogonal transmission of information, and semantic interference elimination mechanism at the receiving end, as well as the end-to-end autoencoder training process.

In step S102, the first user information and the second user information are obtained, the first user information and the second user information are encoded, and the first characteristic symbol and the second characteristic symbol are determined; wherein, the first The user information is user information sent by the central user, and the second user information is user information sent by the edge user. Specifically, the method includes: encoding the first user information and the second user information to determine a first time domain signal and a second time domain signal; and encoding the first time domain signal and the second time domain signal according to the first time domain signal and the second time domain signal. Feature extraction is performed on the first user information and the second user information to determine the first characteristic symbol and the second characteristic symbol. Among _them , the characteristic symbol is recorded as The first characteristic symbol is denoted as X ₁ . Similarly, the edge user is denoted as the second user, the second characteristic symbol is the characteristic symbol of the edge user, and the second characteristic symbol is denoted as X ₂ .

In some optional implementations, as shown in Figure 3, the transmitter includes a pair of encoders designed corresponding to the first user, that is, central user 1, and the second user, that is, edge user 2, that is, the first encoder. and a second encoder, which sequentially compiles the first user information and the second user information through the first encoder and the second encoder, including: source coding, channel coding, modulation mapping, and the result obtained after compilation.

In some optional implementations, the first encoder and the second encoder of NOMA use CNN (Convolutional Neural Network, i.e. convolutional neural network) to perform an operation on the information sources (sources of information) of the first user and the second user. Processing, including source coding, channel coding and modulation mapping processes, generates time domain signals. The pair of coders (i.e. the first code and the second code) respectively encode the information for the center user 1 and the edge user 2 according to the difference in their channel gains. Features are extracted and compressed to generate a modulated signal of characteristic symbols.

In some implementations, as shown in Figure 3, it is assumed that the modulated signal of the characteristic symbol of user m is L is the number of complex characteristic symbols. Assuming that one OFDM symbol has K subcarriers, the characteristic symbols of user m generate a total of N OFDM symbols, where/> Therefore _, the symbol feature X _m is X _m =[x _m,1 ,x _m,2 ,...,x _m,n ,...,x _m,N ], where _{, n,1} ,x _m,n,2 ,...,x _m,n,k ,...x _m,n,K ], x _m,n,k represents the carrier of user m in the nth OFDM symbol Symbol characteristics on k. Symbol characteristics are:

Among them, θ _m is the encoder parameter of user m, and u _m is the message sent by user m. It should be noted that when m is 1, the encoder is the first encoder.

In step S104, determine a first distance value between the first user and the base station, and determine a second distance value between the second user and the base station. Based on the first distance value and the second distance value, determine the first channel gain and the second distance value. Two channel gain. Wherein, the first distance value is the distance between the first user and the base station, the second distance value is the distance between the second user and the base station, and the first distance value is the distance between the first user and the base station. , the second distance value is the distance between the second user and the base station.

Specifically, the method includes: determining the fading path loss of the first user according to the first distance value; determining the fading path loss of the second user according to the second distance value; wherein the fading path loss of the user is large. Scale fading path loss.

In some optional implementations, as shown in Figure 4, the process of determining the first channel gain and the second channel gain for NOMA users to non-orthogonally transmit information specifically includes: in the non-orthogonal multiple access method of this application The distance between the incoming user m and the base station is d _m , and its large-scale fading path loss is:

Therefore, according to It can be determined that the channel gain between user m and the base station is g={g _1,1 ,...,g _m,k ,...}, that is, the first channel gain and the second channel gain can be obtained. Specifically , according to the fading path loss of the first user and the fading path loss of the second user, the first channel gain and the second channel gain are determined by the following formula:

Among them, h _{m, k} is the small-scale fading of user m on carrier k, ρ _m is the fading path loss of user m, m = 1, 2, 3,..., M. The small-scale fading channel model is modeled as having direct radiation The Rician fading of the path, the Rician factor is R.

In step S106, the first user information and the second user information are denormalized according to the first characteristic symbol, the second characteristic symbol, the first channel gain and the second channel gain. Overlay and superimpose, and determine the superimposed signal, specifically including: performing power normalization on the first characteristic symbol and the second characteristic symbol, and determining the first transmission signal of the first user and the second transmission signal of the second user; wherein , the first transmission signal is a transmission signal of the first user on a specific carrier, or the second transmission signal is a transmission signal of the second user on a specific carrier; according to the first transmission signal, the second transmission signal A signal, the first channel gain and the second channel gain are transmitted to determine the superimposed signal.

In some optional implementations, the power normalization includes: obtaining the user priority, the first distance value and the second distance value, and obtaining the user priority, the first distance value or the second distance value according to the user priority and the first distance value or the second distance value. , determine the transmission signal of the first user or the second user on a specific carrier.

In some embodiments, the first symbol characteristic and the second symbol characteristic are subjected to power normalization control and then are sent as transmission signals to the multi-carrier channel. During the power control process, the user is determined according to the user priority and the distance between the user and the base station. The size of user transmission power P _m under each OFDM symbol, where the transmission signal of user m on carrier k of the nth OFDM symbol is as follows:

Among them, x _m,n,k is the symbol characteristics of user m on carrier k of the nth OFDM symbol, P _m is the user transmit power under the OFDM symbol, x _m,n is the user m on the nth OFDM symbol Information.

In some optional implementations, it is assumed that the user's total transmit power constraint is As shown in Figure 4, the signals _zm,n,k of central user 1 and edge user 2 pass through their respective channels and are superimposed at the receiving end. Then, the superimposed signal at the receiver is: y=[y ₁ ,y ₂ ,...,y _n ,...,y _N ], where y _n =[y _n,1 ,y _n,2 ,.. .,y _n,k ,...,y _n,K ], Among them/> It means that the noise w _k obeys the Gaussian distribution with mean 0 and variance σ ² , m is the user, m=1,2,3,...,M, z _m,n,k is the user m in the nth OFDM symbol The transmitted signal of carrier k, g _m,k is the gain signal, which represents the signal gain of user m on carrier k.

In step 108, decoding the superimposed signal, the first channel gain and the second channel gain, and determining the first reconstruction information and the second reconstruction information, including: according to the superimposed signal and the first Channel gain: perform semantic feature recovery and information reconstruction through a convolutional neural network to determine the first reconstructed information; perform signal modulation and power control on the first reconstructed information to determine the receiving end signal corresponding to the first user signal ; According to the receiving end signal, the superimposed signal and the second channel gain, semantic feature recovery and information reconstruction are performed through a convolutional neural network to determine the second reconstruction information.

In some optional implementations, determining the first reconstruction information and the second reconstruction information respectively includes: performing channel equalization processing on the superimposed signal through minimum mean square error, and performing channel equalization processing according to the first channel gain and channel equalization processing. The superimposed signal is used to perform semantic feature recovery and information reconstruction through a convolutional neural network to determine the first reconstructed information. Perform interference elimination on the receiving end signal according to the superimposed signal to determine an interference-free signal; perform channel equalization processing on the interference-free signal through the minimum mean square error, and perform channel equalization processing according to the second channel gain and channel equalization processing. The interference-free signal is used for semantic feature recovery and information reconstruction, and the second reconstructed information is determined.

In some optional implementations, the NOMA user has a semantic interference cancellation mechanism at the receiving end, which specifically includes: performing interference cancellation and information reconstruction on the superimposed signal at the receiving end. As shown in Figure 5, the received signal y is first subjected to MMSE equalization processing. , this signal post-processing algorithm can convert the received signal into a signal that is as close as possible to the transmitted signal. The transmitted signal of central user 1 is first decoded, and the transmitted signal of edge user 2 is treated as interference during signal processing. Therefore, the signal processed by MMSE, that is, the input of the convolutional network of Decoder1 (decoder 1) is:

Among them, p ₁ is the transmit power of each OFDM symbol of the central user 1, H ₁ is the estimated channel gain matrix of the central user 1 in each subcarrier, H ₁ ^H represents the conjugate transpose, I represents the identity matrix, and σ ² is the base station Noise variance at the receiving end.

In some embodiments, user m passes through their respective Decoder and performs semantic feature recovery and information reconstruction through the convolutional neural network in the decoder, and the reconstructed information is determined to be:

in, are the decoder parameters of user m,/> is the signal of user m determined after performing channel equalization processing on the superimposed signal, m is the user, m=1,2,3,...,M. When m is 1, the above reconstruction information is the first reconstruction information. When When m is 2, the above reconstruction information is the second reconstruction information.

In some optional implementations, determining the second reconstruction information includes: after determining the first reconstruction information, using Encoder1 (the decoder corresponding to the central user, that is, the first decoder) in the third encoder pair. Signal modulation, and then power control to obtain a transmit signal of approximately Z ₁ After estimating the channel fading H _1, the signal of the central user 1 at the receiving end is obtained. That is, edge user 2 is interfered by the signal of user 1. Therefore, the interference-free signal at the receiving end of edge user 2 is: Further, channel equalization processing is performed on the interference-free signal according to the minimum mean square error, and channel equalization processing is performed on the receiving end signal y ₂ to determine the following formula: Then the interference-free signal after channel equalization is input to the second decoder for feature recovery and information reconstruction, and the second reconstruction information is determined:

Among them, p ₂ is the transmit power of edge user 2 (that is, the second user), and H ₂ is the estimated channel gain matrix of edge user 2 in each subcarrier. Then the semantic features of edge user 2 are restored through the convolutional network in Decoder2 (second decoder) to obtain the reconstruction information.

In some optional implementations, an end-to-end autoencoder training solution for the user semantic NOMA access system is also included. Specifically, the multi-user semantic information transmission system based on NOMA proposed in the technical solution is an end-to-end training process of joint source channel coding. Compared with the traditional separate source channel coding, the joint source channel using deep learning Coding technology provides new information feature dimensions, can extract semantic features of information, and provides new solutions for improving the performance of information transmission.

From the perspective of end-to-end joint training, the original sending information in this technical solution is u _m and the reconstructed information at the receiving end is The information distortion at the sending and receiving ends of user m is defined as:

Furthermore, considering that users have different priorities and have different requirements for reconstruction quality, for the different requirements λ _m of user m, the priority information distortion is defined as:

In summary, in this system, by optimizing the network parameters θ _m and The problem of minimizing the priority information distortion of NOMA user pairs is modeled as:

in, and/> are the optimal parameters of the codec under channel gain g and channel noise σ ² respectively.

In some optional implementations, in this application, the applicant designed a multi-user semantic information transmission system based on NOMA technology. NOMA users send information to the base station. At the transmitting end, a pair of NOMA encoders are first used to extract features of the central user and edge users to obtain the modulated signal. The modulated signal is then normalized and power controlled, and then the information is sent to the base station for reception. At the end, the receiving end receives the superimposed power signals of the central user and the edge user, and reconstructs the multi-user transmission information through the designed semantic interference elimination mechanism. This application designs a communication system architecture for non-orthogonal multiple access transmission of multi-user semantic information. This is a key point of this patent and should be protected.

Further, in this application, the applicant proposes a multi-user semantic interference cancellation mechanism at the receiving end. At the receiving end, interference elimination and information reconstruction are performed on the superimposed signal. First, the signal sent by the central user is decoded. The received signal needs to be equalized by MMSE (an equalizer, minimizing the received data model) to make the received signal as good as possible. Close to the transmission signal, the signal sent by the edge user is treated as interference during signal processing. After obtaining the information of the central user, the interference of the user can be eliminated at the receiving end, and then the information of the edge user can be decoded and reconstructed. The processing of superimposed signals and the information reconstruction process in the multi-user semantic interference cancellation mechanism proposed in this solution are key points of this patent and should be protected.

Furthermore, the applicant proposed a correlation mechanism between information reconstruction requirements and user power control based on user priority. The priority of the user is determined based on the inherent attributes of the user. In the system proposed by this technical solution, the priority is closely related to the power control of multi-users and the information reconstruction quality requirements of multi-users. Because the stronger the signal, the more accurate the decoding, and the better the reconstruction quality of the information. Therefore, in the design of this solution, the higher the user priority, the greater the transmission power, and the higher the demand for information reconstruction quality, forming a benign promotion mechanism between reconstruction quality and transmission power.

In some embodiments, compared with traditional syntax communication that combines NOMA technology and bit information modulation, this application does not focus on the transmission of syntax information of specific bits, but introduces the semantic dimension of information, extracts low-dimensional information features, and It ensures end-to-end transmission quality while compressing the number of transmission symbols and saving bandwidth resources. At the same time, multi-user bit stream transmission cannot take advantage of the correlation between users. The multi-user semantic information in this application can make use of the correlation characteristics between users. Contributes to interference elimination and better quality information reconstruction at the receiving end.

In some embodiments, compared with the deep learning method based on the model-based learning method, this application studies the communication system architecture with multi-user semantic interference, and designs a corresponding semantic interference elimination mechanism at the receiving end for this purpose. Different processing processes are designed for central users and edge users respectively to reduce interference and separate users' respective semantic information for reconstruction; this application uses a pair of NOMA encoders to fully utilize the correlation characteristics between users for collaborative encoding, improving user access. input, and the utilization of correlation features also improves the quality of information reconstruction.

In summary, the present invention proposes to realize multi-user semantic information transmission based on NOMA technology, fully explore the correlation of semantic information between users, extract effective semantic features, increase the number of user accesses, and improve the quality of information reconstruction.

It can be seen from the above that the multi-user semantic information transmission method based on non-orthogonal multiple access provided by this application includes: obtaining the first user information and the second user information, and comparing the first user information and the so-called first user information. The second user information is encoded to determine the first characteristic symbol and the second characteristic symbol; wherein the first user information is central user information, and the second user information is edge user information; the relationship between the first user and the base station is determined. The first distance value, the second distance value between the second user and the base station, determine the first channel gain and the second channel gain according to the first distance value and the second distance value; according to the first characteristic symbol, The second characteristic symbol, the first channel gain and the second channel gain perform non-orthogonal superposition on the first user information and the second user information to determine a superimposed signal; according to the superimposed signal , decoding the first channel gain and the second channel gain to determine the first reconstruction information and the second reconstruction information. This application designs different processing procedures for central users and edge users respectively to reduce interference and separate the users' respective semantic information for reconstruction. This proposal improves the information carrying capacity under a specific bandwidth through data dimensionality reduction and compression, also increases the number of user accesses, and uses the correlation characteristics of multiple users to improve the quality of information reconstruction.

It should be noted that the method in the embodiment of the present application can be executed by a single device, such as a computer or server. The method of this embodiment can also be applied in a distributed scenario, and is completed by multiple devices cooperating with each other. In this distributed scenario, one of the multiple devices can only execute one or more steps in the method of the embodiment of the present application, and the multiple devices will interact with each other to complete all the steps. method described.

It should be noted that some embodiments of the present application have been described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the above-described embodiments and still achieve the desired results. Additionally, the processes depicted in the figures do not necessarily require the specific order shown, or sequential order, to achieve desirable results. Multitasking and parallel processing are also possible or may be advantageous in certain implementations.

Based on the same inventive concept and corresponding to any of the above embodiment methods, this application also provides a multi-user semantic information transmission system based on non-orthogonal multiple access.

Referring to Figure 6, the multi-user semantic information transmission system based on non-orthogonal multiple access includes:

Coding module, transmitter, multi-carrier channel, receiver and interference cancellation module;

The transmitter includes the encoding module, which includes a first encoder, a second encoder, and a power control module. The transmitter encodes the first user information and the second user information through the encoding module. , determine the first characteristic symbol and the second characteristic symbol corresponding to the first user information and the second user information, perform power normalization on the first characteristic symbol and the second characteristic symbol, and The normalized first characteristic symbol and the second characteristic symbol are sent to the multi-carrier channel as transmission signals;

The multi-carrier channel is used to perform non-orthogonal transmission of the first characteristic symbol and the second characteristic symbol, and determine a first distance value between the first user and the base station, and a second distance value between the second user and the base station, According to the first distance value and the second distance value, a first channel gain and a second channel gain are determined, and the first channel gain and the second channel gain are sent to the receiver and the Interference cancellation module;

The receiver includes the interference cancellation module, which includes a first decoder, a second decoder, and a third encoder. The receiver is configured to receive the first channel gain and the The second channel gain is to superpose the first user information and the second user information according to the first characteristic symbol and the second characteristic symbol, the first channel gain and the second channel gain. , determine the superimposed signal, perform interference elimination and information reconstruction on the superimposed signal, and determine the reconstructed information.

For the convenience of description, when describing the above device, the functions are divided into various modules and described separately. Of course, when implementing this application, the functions of each module can be implemented in the same or multiple software and/or hardware.

The devices of the above embodiments are used to implement the corresponding multi-user semantic information transmission method based on non-orthogonal multiple access in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be described again here.

Based on the same inventive concept, corresponding to any of the above embodiments, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor When the program is executed, the multi-user semantic information transmission method based on non-orthogonal multiple access as described in any of the above embodiments is implemented.

FIG. 7 shows a more specific hardware structure diagram of an electronic device provided by this embodiment. The device may include: a processor 1010, a memory 1020, an input/output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input/output interface 1030 and the communication interface 1040 implement communication connections between each other within the device through the bus 1050.

The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit, central processing unit), a microprocessor, an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), or one or more integrated circuits, and is used to execute related tasks. program to implement the technical solutions provided by the embodiments of this specification.

The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store operating systems and other application programs. When implementing the technical solutions provided by the embodiments of this specification through software or firmware, the relevant program codes are stored in the memory 1020 and called and executed by the processor 1010 .

The input/output interface 1030 is used to connect the input/output module to realize information input and output. The input/output/module can be configured in the device as a component (not shown in the figure), or can be externally connected to the device to provide corresponding functions. Input devices can include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices can include monitors, speakers, vibrators, indicator lights, etc.

The communication interface 1040 is used to connect a communication module (not shown in the figure) to realize communication interaction between this device and other devices. The communication module can realize communication through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

Bus 1050 includes a path that carries information between various components of the device (eg, processor 1010, memory 1020, input/output interface 1030, and communication interface 1040).

It should be noted that although the above device only shows the processor 1010, the memory 1020, the input/output interface 1030, the communication interface 1040 and the bus 1050, during specific implementation, the device may also include necessary components for normal operation. Other components. In addition, those skilled in the art can understand that the above-mentioned device may only include components necessary to implement the embodiments of this specification, and does not necessarily include all components shown in the drawings.

The electronic devices of the above embodiments are used to implement the corresponding multi-user semantic information transmission method based on non-orthogonal multiple access in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be described again here.

Based on the same inventive concept, corresponding to any of the above embodiment methods, the present application also provides a non-transitory computer-readable storage medium, the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions use To enable the computer to execute the multi-user semantic information transmission method based on non-orthogonal multiple access as described in any of the above embodiments.

The computer-readable media in this embodiment include permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information may be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), and read-only memory. (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, Magnetic tape cassettes, tape magnetic disk storage or other magnetic storage devices or any other non-transmission medium can be used to store information that can be accessed by a computing device.

The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the multi-user semantic information transmission method based on non-orthogonal multiple access as described in any of the above embodiments, and have the benefits of the corresponding method embodiments. The effect will not be described again here.

Based on the same inventive concept, corresponding to the multi-user semantic information transmission based on non-orthogonal multiple access described in any of the above embodiments, the present disclosure also provides a computer program product, which includes computer program instructions. In some embodiments, the computer program instructions may be executed by one or more processors of a computer to cause the computer and/or the processor to perform the color correction method. Corresponding to the execution subject corresponding to each step in each embodiment of the multi-user semantic information transmission method based on non-orthogonal multiple access, the processor that executes the corresponding step may belong to the corresponding execution subject.

The computer program product of the above embodiments is used to cause the computer and/or the processor to execute the multi-user semantic information transmission method based on non-orthogonal multiple access as described in any of the above embodiments, and has a corresponding method The beneficial effects of the embodiment will not be described again here.

Those of ordinary skill in the art should understand that the discussion of any above embodiments is only illustrative, and is not intended to imply that the scope of the present application (including the claims) is limited to these examples; under the spirit of the present application, the above embodiments or Technical features in different embodiments can also be combined, steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of the present application as described above, which are not provided in detail for the sake of simplicity.

In addition, to simplify illustration and discussion, and so as not to obscure the embodiments of the present application, well-known power supplies/power supplies with integrated circuit (IC) chips and other components may or may not be shown in the provided figures. Ground connection. Furthermore, devices may be shown in block diagram form in order to avoid obscuring the embodiments of the present application, and this also takes into account the fact that details regarding the implementation of these block diagram devices are highly dependent on the implementation of the embodiments of the present application. platform (i.e., these details should be well within the understanding of those skilled in the art). Where specific details (eg, circuits) are set forth to describe exemplary embodiments of the present application, it will be apparent to those skilled in the art that construction may be accomplished without these specific details or with changes in these specific details. The embodiments of this application are implemented below. Accordingly, these descriptions should be considered illustrative rather than restrictive.

Although the present application has been described in conjunction with specific embodiments thereof, many alternatives, modifications and variations of these embodiments will be apparent to those of ordinary skill in the art from the foregoing description. For example, other memory architectures such as dynamic RAM (DRAM) may use the discussed embodiments.

The present embodiments are intended to embrace all such alternatives, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of this application shall be included in the protection scope of this application.

Claims (10)

1. A multi-user semantic information transmission method based on non-orthogonal multiple access, which is characterized by including: Obtain the first user information and the second user information, encode the first user information and the second user information, and determine the first characteristic symbol and the second characteristic symbol; wherein the first user information is the central user The second user information is the user information of the edge user; Determine a first distance value between the first user and the base station, a second distance value between the second user and the base station, and determine the first channel gain and the second channel gain based on the first distance value and the second distance value; According to the first characteristic symbol, the second characteristic symbol, the first channel gain and the second channel gain, perform non-orthogonal superposition on the first user information and the second user information to determine Overlay signals; The superimposed signal, the first channel gain and the second channel gain are decoded to determine first reconstruction information and second reconstruction information. 2. The method according to claim 1, wherein encoding the first user information and the second user information and determining the first characteristic symbol and the second characteristic symbol includes: Encode the first user information and the second user information to determine a first time domain signal and a second time domain signal; Feature extraction is performed on the first user information and the second user information according to the first time domain signal and the second time domain signal, and the first characteristic symbol and the second characteristic symbol are determined. 3. The method according to claim 1, characterized in that, the method further comprises: Perform power normalization on the first characteristic symbol and the second characteristic symbol to determine the first transmission signal of the first user and the second transmission signal of the second user; wherein the first transmission signal is the first transmission signal. The user transmits a signal on a specific carrier, and the second transmission signal is a signal transmitted by the second user on a specific carrier; The superposition signal is determined based on the first transmission signal, the second transmission signal, the first channel gain and the second channel gain. 4. The method of claim 3, wherein the power normalization includes: Obtain the user priority, the first distance value and the second distance value, and determine the location of the first user or the second user on a specific carrier based on the user priority and the first distance value or the second distance value. Send a signal. 5. The method according to claim 4, wherein the first distance value is the distance between the first user and the base station, and the second distance value is the distance between the second user and the base station. 6. The method according to claim 5, characterized in that the first channel gain is the channel gain between the first user and the base station, and the second channel gain is the channel gain between the second user and the base station. ; Determining the first channel gain and the second channel gain according to the first distance value and the second distance value includes: Determine the fading path loss of the first user according to the first distance value; Determine the fading path loss of the second user according to the second distance value; According to the fading path loss of the first user and the fading path loss of the second user, the first channel gain and the second channel gain are determined by the following formula: Among them, h _m,k is the small-scale fading of user m on carrier k, ρ _m is the fading path loss of user m, m=1,2,3,...,M. 7. The method according to claim 1, wherein the decoding includes semantic feature recovery and information reconstruction; Decoding according to the superimposed signal, the first channel gain and the second channel gain to determine the first reconstruction information and the second reconstruction information includes: According to the superimposed signal and the first channel gain, perform semantic feature recovery and information reconstruction through a convolutional neural network, and determine the first reconstruction information; Perform signal modulation and power control on the first reconstructed information to determine the receiving end signal corresponding to the first user signal; According to the receiving end signal, the superimposed signal and the second channel gain, semantic feature recovery and information reconstruction are performed through a convolutional neural network to determine the second reconstruction information. 8. The method according to claim 7, characterized in that the method further includes: The superimposed signal is subjected to channel equalization processing through minimum mean square error, and semantic feature recovery and information reconstruction are performed through a convolutional neural network according to the first channel gain and the superimposed signal after channel equalization processing, and the first channel is determined. 1. Reconstruction information. 9. The method according to claim 7, characterized in that, based on the receiving end signal, the superimposed signal and the second channel gain, semantic feature recovery and information reconstruction are performed through a convolutional neural network to determine The second reconstruction information includes: Perform interference elimination on the receiving end signal according to the superimposed signal to determine that there is no interference signal; Perform channel equalization processing on the interference-free signal through minimum mean square error, perform semantic feature recovery and information reconstruction based on the second channel gain and the interference-free signal after channel equalization processing, and determine the second reconstructed information . 10. A multi-user semantic information transmission system based on non-orthogonal multiple access, characterized by including: a coding module, a transmitter, a multi-carrier channel, a receiver and an interference cancellation module; The transmitter includes the encoding module, which includes a first encoder, a second encoder, and a power control module. The transmitter encodes the first user information and the second user information through the encoding module. , determine the first characteristic symbol and the second characteristic symbol corresponding to the first user information and the second user information, perform power normalization on the first characteristic symbol and the second characteristic symbol, and The normalized first characteristic symbol and the second characteristic symbol are sent to the multi-carrier channel as transmission signals; The multi-carrier channel is used to perform non-orthogonal transmission of the first characteristic symbol and the second characteristic symbol, and determine a first distance value between the first user and the base station, and a second distance value between the second user and the base station, According to the first distance value and the second distance value, a first channel gain and a second channel gain are determined, and the first channel gain and the second channel gain are sent to the receiver and the Interference cancellation module; The receiver includes the interference cancellation module, which includes a first decoder, a second decoder, and a third encoder. The receiver is configured to receive the first channel gain and the The second channel gain is to superpose the first user information and the second user information according to the first characteristic symbol and the second characteristic symbol, the first channel gain and the second channel gain. , determine the superimposed signal, perform interference elimination and information reconstruction on the superimposed signal, and determine the reconstructed information.