Semantic communication method and apparatus, system, electronic device, computer storage medium

Abstract

The present disclosure provides a semantic communication method and device, relates to the technical field of communication, in particular to the technical fields of large model, deep learning, diversity processing and the like. The specific implementation scheme is: obtaining source data; determining a semantic code symbol sequence and an importance ranking sequence number of a semantic code symbol in the semantic code symbol sequence based on the source data; obtaining a target code symbol sequence based on the importance ranking sequence number and the semantic code symbol sequence; and transmitting the target code symbol sequence to a receiving end through different independent fading paths, so that the receiving end obtains target data corresponding to the source data.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of computers, in particular to the technical field of large models, deep learning, diversity processing, and the like, and more particularly to a semantic communication method and apparatus, a semantic communication system, an electronic device, a computer readable storage medium, and a computer program product. BACKGROUND

[0002] Traditional communication systems treat encoded symbols as equal when transmitting signals, and focus on the accurate transmission of encoded symbols from the transmitter to the receiver. Task-oriented semantic communication mainly focuses on the semantic information carried in the encoded symbols.

[0003] The goal of a semantic communication system is to transmit semantic information, but semantic errors occur during signal transmission. In particular, under low signal-to-noise ratio conditions, noise has a large interference on semantic symbols, which seriously affects the semantic recovery performance of the receiving end. SUMMARY

[0004] The present disclosure provides a semantic communication method and apparatus, a semantic communication system, an electronic device, a computer readable storage medium, and a computer program product.

[0005] According to a first aspect, a semantic communication method is provided, the method comprising: obtaining source data; determining a sequence of semantic encoding symbols and an importance ranking sequence number of the semantic encoding symbols in the sequence of semantic encoding symbols based on the source data; obtaining a target sequence of encoding symbols based on the importance ranking sequence number and the sequence of semantic encoding symbols; and transmitting the target sequence of encoding symbols to a receiving end through different independent fading paths, so that the receiving end obtains target data corresponding to the source data.

[0006] According to a second aspect, another semantic communication method is provided, the method comprising: receiving a target sequence of encoding symbols transmitted through different independent fading paths; merging first semantic encoding symbols having the same semantics in the target sequence of encoding symbols; obtaining an intermediate sequence of encoding symbols based on the first semantic encoding symbols and second semantic encoding symbols in the sequence of semantic encoding symbols that do not have the same semantics; and obtaining target data corresponding to the source data based on the intermediate sequence of encoding symbols and the first semantic encoding symbols.

[0007] According to a third aspect, a semantic communication apparatus is provided, the apparatus comprising: an obtaining unit configured to obtain source data; a determining unit configured to determine a sequence of semantic encoding symbols and an importance ranking sequence number of the semantic encoding symbols in the sequence of semantic encoding symbols based on the source data; an obtaining unit configured to obtain a target sequence of encoding symbols based on the importance ranking sequence number and the sequence of semantic encoding symbols; and a transmitting unit configured to transmit the target sequence of encoding symbols to a receiving end through different independent fading paths, so that the receiving end obtains target data corresponding to the source data.

[0008] According to a fourth aspect, there is provided another semantic communication apparatus, comprising: a receiving unit configured to receive a target coded symbol sequence transmitted through different independent fading paths; a merging unit configured to merge semantic coded symbols with the same semantics in the target semantic coded symbol sequence; a combining unit configured to obtain an intermediate coded symbol sequence based on a first semantic coded symbol and a second semantic coded symbol in the target coded symbol sequence without the same semantics; and a restoring unit configured to obtain target data corresponding to the source data based on the intermediate coded symbol sequence and the first semantic coded symbol.

[0009] According to a fifth aspect, there is provided a semantic communication system, comprising: a receiving end and a sending end; the sending end is configured to obtain source data; determine a semantic coded symbol sequence and an importance order sequence of semantic coded symbols in the semantic coded symbol sequence based on the source data; obtain a target coded symbol sequence based on the importance order sequence and the semantic coded symbol sequence; and transmit the target coded symbol sequence through different independent fading paths to the receiving end; and the receiving end is configured to receive the target coded symbol sequence transmitted through different independent fading paths; merge first semantic coded symbols with the same semantics in the target coded symbol sequence; obtain an intermediate coded symbol sequence based on the first semantic coded symbols and second semantic coded symbols in the target coded symbol sequence without the same semantics; and obtain target data corresponding to the source data based on the intermediate coded symbol sequence and the first semantic coded symbol.

[0010] According to a sixth aspect, there is provided an electronic device, comprising: at least one processor; and a memory connected with the at least one processor in communication, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method described in any implementation manner of the first aspect or the second aspect.

[0011] According to a seventh aspect, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method described in any implementation manner of the first aspect or the second aspect.

[0012] The semantic communication method and device provided by the embodiments of the present disclosure first acquire source data; secondly, based on the source data, determine a semantic code symbol sequence and an importance ranking sequence number of the semantic code symbols in the semantic code symbol sequence; thirdly, based on the importance ranking sequence number and the semantic code symbol sequence, obtain a target code symbol sequence; and finally, transmit the target code symbol sequence to a receiving end through different independent fading paths, so that the receiving end obtains target data corresponding to the source data. Thus, by combining the importance characteristics of the semantic code symbols, reordering and resource allocation are performed on the semantic code symbols to determine the target code symbol sequence, and the target code symbol sequence is transmitted, thereby preferentially guaranteeing the reliable transmission of key information, improving the recovery performance of information, and being particularly suitable for use in an environment with poor channel conditions, and significantly improving the reliability and accuracy of communication.

[0013] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0014] The accompanying drawings are used to better understand the present scheme and do not limit the present disclosure. Among them:

[0015] Figure 1 is a flowchart of one embodiment of the semantic communication method according to the present disclosure;

[0016] Figure 2 is a structural schematic diagram of a system corresponding to the semantic communication method of the present disclosure;

[0017] Figure 3 is a structural schematic diagram of a sending end processing and transmitting a target code symbol sequence in the present disclosure;

[0018] Figure 4 is a flowchart of another embodiment of the semantic communication method according to the present disclosure;

[0019] Figure 5 is a structural schematic diagram of a receiving end receiving and processing a target code symbol sequence in the present disclosure;

[0020] Figure 6 is a structural schematic diagram of one embodiment of the semantic communication device according to the present disclosure;

[0021] Figure 7 is a structural schematic diagram of one embodiment of the semantic communication device according to the present disclosure;

[0022] Figure 8 is a structural schematic diagram of one embodiment of the semantic communication system according to the present disclosure;

[0023] Figure 9 is a block diagram of an electronic device for implementing the semantic communication method of the embodiments of the present disclosure. DETAILED DESCRIPTION

[0024] Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, which include various details of the embodiments of the present disclosure to assist in understanding, which should be considered in a descriptive sense only. Thus, it will be apparent to one of ordinary skill in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Also, for the sake of brevity and clarity, descriptions of well-known functions and constructions are omitted from the following description.

[0025] The diversity technique in wireless communication involves transmitting the same data on independent fading paths, and since the probability of independent fading paths experiencing deep fading at the same time is small, appropriate combining can reduce the fading degree of the received signal.

[0026] In a conventional semantic communication system, each semantic encoding symbol after semantic encoding has different contribution degrees to downstream tasks, and has significant inequality, and semantic errors occur in the process of diversity transmission. In view of this defect, the present disclosure proposes a semantic communication method, Figure 1 Flow 100 according to one embodiment of the semantic communication method of the present disclosure is shown, and the above semantic communication method comprises the following steps:

[0027] Step 101, obtaining source data.

[0028] In this embodiment, the source data is a data source to be transmitted to the receiving end. The source data can be in any form, such as images, texts, video point clouds, etc. In particular, the source data can be channel state information (CSI), which describes the fading factor of the signal on each transmission path, i.e., the value of each element in the channel gain matrix, such as signal scattering (Scattering), environmental fading (Fading, multipath fading or shadowing fading), distance attenuation (Power Decay of Distance), etc.

[0029] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of the source data are performed after authorization, in accordance with relevant laws and regulations. The information related to the user in the source data is obtained after the user's permission.

[0030] Step 102, based on the source data, determining a sequence of semantic encoding symbols and an importance ranking sequence number of the semantic encoding symbols in the sequence of semantic encoding symbols.

[0031] In the embodiment, the sequence of semantic coding symbols is a sequence formed by a plurality of semantic coding symbols, the semantic coding symbol is a symbol obtained after semantic coding of a single data in the source data, and the step 102 includes: performing data preprocessing on the source data to obtain processed data; performing semantic coding on the processed data to obtain a sequence of semantic coding symbols including a plurality of semantic coding symbols; and identifying the importance of each semantic coding symbol in the entire sequence to obtain an importance ranking sequence number of each semantic coding symbol.

[0032] In the embodiment, the importance ranking sequence number is used to represent the semantic importance of the semantic coding symbol, and the importance ranking sequence number can be obtained by various means, for example, directly calculating the information entropy of each semantic coding symbol, ranking the information entropy of all semantic coding symbols in the sequence of semantic coding symbols, and obtaining the sequence number of each semantic coding symbol as the importance ranking sequence number.

[0033] In the embodiment, the execution subject on which the semantic communication method runs is a sending end UE shown in FIG. 1, the source data Hd is coded by an encoder in the sending end UE to obtain a sequence of semantic coding symbols, and the target coding symbol sequence is obtained by ranking the sequence of semantic coding symbols according to the importance. Figure 2

[0034] Step 103: obtaining a target coding symbol sequence based on the importance ranking sequence number and the sequence of semantic coding symbols.

[0035] In the embodiment, the target coding symbol sequence is a symbol sequence to be transmitted to a receiving end (which can be directly transmitted to the receiving end through a channel), and the target data representing the semantics of the source data can be obtained by decoding the target coding symbol sequence, wherein the target data can be the entire source data or data having only part of the source data.

[0036] In the embodiment, each semantic coding symbol in the sequence of semantic coding symbols is sorted in descending order of importance according to the importance corresponding to the importance ranking sequence number, and the last set of semantic coding symbols in the sorted sequence of semantic coding symbols is removed to obtain the target coding symbol sequence.

[0037] In the embodiment, the sequence number value of the importance ranking sequence number can represent the importance degree of each semantic coding symbol, and optionally, when the importance ranking sequence number is larger, the importance degree is stronger, the importance ranking sequence number is sorted in descending order, the last set of importance ranking sequence numbers is removed, and the semantic coding symbols corresponding to the sorted importance ranking sequence numbers are used as the target coding sequence.

[0038] ​Step 104, transmitting the target coded symbol sequence to the receiving end through different independent fading paths, so that the receiving end obtains the target data corresponding to the source data.

[0039] As shown in Figure 2 the receiving end BS receives the target coded symbol sequence transmitted through different independent fading paths, processes the target coded symbol sequence, and obtains the target data Md.

[0040] In this embodiment, each target coded symbol in the target coded symbol sequence is transmitted through different independent fading paths, and when there are repeated target coded symbols in the target coded symbol sequence, the effect of diversity transmission can be achieved.

[0041] In this embodiment, the diversity technology in wireless communication involves transmitting the same data on independent fading paths. Since the probability of independent paths experiencing deep fading at the same time is small, the fading degree of the received signal can be reduced after appropriate combination. Diversity is a method of processing the received fading characteristics independently to reduce the fluctuation of the signal level. Diversity refers to the transmission of multiple statistically independent fading signals carrying the same information and the reception of the multiple statistically independent fading signals to reduce the impact of fading.

[0042] In this embodiment, according to the principle of signal theory, if other attenuated original transmitted signal copies are provided to the receiver, it is helpful for the correct decision of the received signal. This method of improving the correct decision rate of the received signal by providing multiple copies of the transmitted signal is called diversity. Diversity technology is used to compensate for the loss of fading channels. It usually uses the characteristics of the independence between independent samples of the same signal in the wireless propagation environment, and uses certain signal combination techniques to improve the received signal.

[0043] In this embodiment, when there are multiple semantic coded symbols with the same semantics in the target coded symbol sequence, transmitting the target coded symbols in the target coded symbol sequence through multiple different independent fading paths can achieve the same effect of signal diversity transmission. Specifically, diversity technology involves transmitting the same data on independent fading paths (this is the concept of diversity, which requires independent fading paths and the same data). In this embodiment, as shown in Figure 3The target coding symbol sequence Sd is shown in the figure, each block represents a symbol, the darker the color, the more important the degree, which is special here, because each symbol is transmitted through an OFDM subcarrier, so each symbol is transmitted on an independent fading path, so the important symbols at the beginning and end constitute the concept of diversity transmission (but in fact each symbol is transmitted in the same way). Since diversity transmission can resist fading, important semantic coding symbols in the target coding symbol sequence Sd transmitted by diversity transmission can ensure the transmission of important semantic symbols and thus ensure the overall transmission quality.

[0044] In this embodiment, based on the distribution of target coding symbols in the target coding symbol sequence, the receiving end can achieve the purpose of protecting important semantic coding symbols regardless of the transmission method or technology used in the subsequent. When the target coding symbol sequence uses diversity technology transmission, the diversity technology can be divided into spatial diversity, frequency diversity, time diversity, etc. according to the different communication resources used. Generally speaking, diversity in communication refers to spatial diversity, that is, the same information is transmitted between different antennas to increase reliability, which is usually used in MIMO (Multiple-Input Multiple-Output) systems.

[0045] In this embodiment, the target coding symbol sequence can be transmitted to the receiving end using frequency diversity, which specifically uses multiple different frequency subcarriers to transmit the same information, thereby reducing the impact of frequency selective fading, that is, the independent fading path is the frequency subcarrier. The frequency diversity technology here uses multiple different frequency subcarriers to transmit the same information, thereby reducing the impact of frequency selective fading. For specific implementation, refer to Figure 3 That is, the semantic coding symbol sequence S after encoding by the encoder E in the sending end, selectively discarding the non-important semantic symbols at the end of the sequence, and re-inserting the semantic coding symbols corresponding to the most important subsequence (such as Figure 3 The content in the curly braces shown), to obtain the target coding symbol sequence Sd, which is mapped by the OFDM (Orthogonal Frequency Division Multiplexing) transmitter of the OFDM (Orthogonal Frequency Division Multiplexing) transmitter to obtain the OFDM symbol Os. The OFDM subcarriers vacated during the transmission of the OFDM symbol Os will be reallocated for frequency diversity transmission of the important semantic symbols at the beginning of the sequence.

[0046] In this embodiment, the receiving end is a terminal that communicates with the execution subject on which the semantic communication method runs, that is, the execution subject and the receiving end can transmit information to each other through a channel.

[0047] In this embodiment, the target data corresponding to the source data is the data obtained after the receiving end restores the target code symbol sequence, which can be the source data or part of the source data. The target data has the same importance as the source data in the data transmission task and can achieve the same effect as the source data.

[0048] The semantic communication method provided by the embodiment of the present disclosure first acquires source data, then determines a semantic code symbol sequence and an importance ranking sequence number of a semantic code symbol in the semantic code symbol sequence based on the source data, further obtains a target code symbol sequence based on the importance ranking sequence number and the semantic code symbol sequence, and finally transmits the target code symbol sequence to the receiving end through different independent fading paths, so that the receiving end obtains target data corresponding to the source data. Therefore, by combining the importance characteristics of the semantic code symbol, reordering and resource allocation are performed on the semantic code symbol to determine the target code symbol sequence, and the target code symbol sequence is transmitted in diversity, thereby preferentially guaranteeing the reliable transmission of key information, improving the recovery performance of information, and being particularly suitable for environments with poor channel conditions, and significantly improving the reliability and accuracy of communication.

[0049] In some optional implementations of the present disclosure, the above determining, based on the source data, the semantic code symbol sequence and the importance ranking sequence number of the semantic code symbol in the semantic code symbol sequence includes: determining a data type based on the source data; determining a semantic space position model based on the data type; inputting the source data into the semantic space position model to obtain the semantic code symbol sequence output by the semantic space position model and the importance ranking sequence number of the semantic code symbol in the semantic code symbol sequence.

[0050] In this optional implementation, the data type includes text, image, video, point cloud, etc. The purpose of determining the data type is to use the semantic encoder model corresponding to the source of the corresponding type for semantic encoding and use the semantic decoder model corresponding to the source of the corresponding type for semantic decoding.

[0051] In this optional implementation, the semantic space position model can include a model for processing source data of multiple data types, and the semantic space position model for processing source data of a certain data type can be directly obtained through the data type of the source data.

[0052] In this optional implementation, the semantic spatial position model is obtained by training a semantic encoder model based on a mask-based adaptive deep coding training framework in an end-to-end manner. After the semantic encoder model trained under such an adaptive deep coding training framework performs semantic coding on the source, the semantic code symbols formed naturally present a phenomenon of importance descending order in spatial position. Therefore, the semantic spatial position model can directly obtain the semantic code symbol sequence and the importance ranking sequence number of the semantic code symbols in the semantic code symbol sequence.

[0053] The method for determining the semantic code symbol sequence and the importance of the semantic code symbol sequence provided by this optional implementation is based on source data to determine the data type, determine the semantic spatial position model based on the data type, and input the source data into the semantic spatial position model to obtain the semantic code symbol sequence output by the semantic spatial position model and the importance ranking sequence number of the semantic code symbols in the semantic code symbol sequence. The semantic code symbol and the importance ranking sequence number are directly obtained by the semantic spatial position model with a defined spatial distribution, which is simple and convenient to implement and improves the efficiency of obtaining the semantic code symbol sequence and the importance of the semantic code symbol sequence.

[0054] In some optional implementations of the present disclosure, the above method for determining the semantic code symbol sequence and the importance ranking sequence number of the semantic code symbols in the semantic code symbol sequence based on source data includes determining the data type based on the source data, determining the semantic encoder based on the data type, inputting the source data into the semantic encoder to obtain the semantic code symbol sequence output by the semantic encoder, and inputting the semantic code symbol sequence into a pre-set importance ranking model to obtain the importance ranking sequence number of the semantic code symbols in the semantic code symbol sequence output by the importance ranking model.

[0055] In this optional implementation, the semantic encoder can include encoders for semantic coding of source data of multiple data types. The encoder for processing source data of a certain data type can be directly obtained by the data type of the source data.

[0056] In this optional implementation, the semantic encoder can customize different coding strategies according to different scenarios, such as sending the most accurate message, the easiest generated message, or the message required by the receiver, etc. Similar to the transmitter of the communication principle, the semantic encoder can process how to reduce the redundancy in the message (source coding) and how to improve the reliability of the transmission (channel coding).

[0057] In this optional implementation, the semantic code symbol sequence is input into a pre-set importance ranking model. The importance ranking model can be an entropy model, a gradient model, or other neural networks for measuring semantic importance.

[0058] The method for determining the semantic code symbol sequence and the importance of the semantic code symbol sequence provided by the optional implementation mode determines the data type based on the source data, determines the semantic encoder based on the data type, inputs the source data into the semantic encoder to obtain the semantic code symbol sequence of the semantic output output by the semantic encoder, and inputs the semantic code symbol sequence into the pre-set importance ranking model to obtain the importance ranking sequence number of the semantic code symbol in the semantic code symbol sequence output by the importance ranking model, thereby providing a reliable implementation mode for obtaining the semantic code symbol sequence and the importance of the semantic code symbol sequence.

[0059] In some optional implementation modes of the present disclosure, the above method for obtaining the target code symbol sequence based on the importance ranking sequence number and the semantic code symbol sequence comprises: determining a less important sub-sequence number based on the importance ranking sequence number; removing the semantic code symbol corresponding to the less important sub-sequence number in the semantic code symbol sequence to obtain an intermediate code symbol sequence; determining a most important sub-sequence number having the same number of bits as the less important sub-sequence number based on the importance ranking sequence number; and supplementing the semantic code symbol corresponding to the most important sub-sequence number after the intermediate code symbol sequence to obtain the target code symbol sequence.

[0060] In the optional implementation mode, the less important sub-sequence number is the sequence number of the semantic code symbol with poor importance in the importance ranking sequence number, and the number of less important sub-sequence numbers can be multiple, each less important sub-sequence number corresponds to a semantic code symbol, and removing the semantic code symbol corresponding to the less important sub-sequence number in the semantic code symbol sequence can make the target code symbol sequence obtained only have the semantic code symbol with high importance.

[0061] In the optional implementation mode, the most important sub-sequence number is the sequence number of the most important semantic code symbol in the importance ranking sequence number, and the number of most important sub-sequence numbers can be one or multiple, each most important sub-sequence number corresponds to a semantic code symbol.

[0062] In the optional implementation mode, the above method for determining the most important sub-sequence number having the same number of bits as the less important sub-sequence number based on the importance ranking sequence number comprises: determining the number of bits of the less important sub-sequence number; and determining the most important sub-sequence number based on the importance ranking of the importance ranking sequence number. For example, if the number of bits of the less important sub-sequence number is 3 bits, the sequence number with the highest importance in the importance ranking sequence number is taken as the most important sub-sequence.

[0063] In the optional implementation, the target coding symbol sequence is a sequence with the same length as the semantic coding symbol sequence after semantic coding symbol removal and supplementing, and the removed and supplemented semantic coding symbols in the semantic coding symbol sequence can be determined based on requirements. Generally, semantic coding symbols that have little impact on semantic communication transmission can be removed, and semantic coding symbols that have a great impact on semantic communication transmission need to be supplemented again. The semantic coding symbols in the target coding symbol sequence are also coding symbols that need to be transmitted by the semantic communication method. Since the target coding symbol sequence does not have semantic coding symbols with poor importance, the semantic communication transmission can transmit semantic coding symbols with high importance. Moreover, since the target coding symbol sequence has multiple semantic coding symbols with high importance, when the target coding symbol sequence is transmitted through different fading paths, the effect of diversity transmission of semantic coding symbols with the same semantics can be achieved.

[0064] The method for obtaining the target coding symbol sequence provided by the optional implementation includes determining a less important subsequence number based on the importance ranking sequence number, removing the semantic coding symbols corresponding to the less important subsequence number in the semantic coding symbol sequence to obtain an intermediate coding symbol sequence, determining a most important subsequence number with the same number of bits as the less important subsequence number based on the importance ranking sequence number, and supplementing the semantic coding symbols corresponding to the most important subsequence number after the intermediate coding symbol sequence to obtain the target coding symbol sequence. This provides a reliable implementation means for obtaining the target coding symbol sequence.

[0065] Optionally, the method for obtaining the target coding symbol sequence based on the importance ranking sequence number and the semantic coding symbol sequence includes determining a high importance subsequence number based on the importance ranking sequence number, selecting the semantic coding symbols corresponding to the high importance subsequence number in the semantic coding symbol sequence as the intermediate coding symbol sequence, and determining the most important semantic coding symbol based on the number of bits of the sequence numbers other than the high importance subsequence number in the importance ranking sequence number, and supplementing the most important semantic coding symbol after the intermediate coding symbol sequence to obtain the target coding symbol sequence. The high importance subsequence number is the sequence number of the semantic coding symbol with high importance.

[0066] In some optional implementations of the present disclosure, the method for determining the less important subsequence number based on the importance ranking sequence number includes determining the proportion of non-important coding symbols based on the diversity ratio, and determining the less important subsequence number based on the importance ranking sequence number and the proportion.

[0067] In the optional implementation, the diversity ratio is a proportional value set by the semantic diversity module and can be adaptively adjusted according to the current channel condition. The diversity ratio is used to control the number of important symbols in the diversity transmission. The value of the diversity ratio should be larger when the channel condition is worse, so as to ensure the transmission of more important symbols. When the channel condition improves, the value of the diversity ratio decreases. In the optional implementation, the proportion is the proportion of the non-important coded symbols in the semantic coded symbol sequence, that is, the proportion is equal to the ratio of the number of the non-important coded symbols to the number of the semantic coded symbols in the semantic coded symbol sequence. When the diversity order is larger, the proportion of the non-important coded symbols in the semantic coded symbol sequence is smaller. When the diversity order is smaller, the proportion of the non-important coded symbols in the semantic coded symbol sequence is larger. The corresponding relationship between the diversity order and the proportion can be obtained by querying the calibration table.

[0068] In the optional implementation, the method for determining the less important sub-sequence based on the importance ranking sequence number and the proportion includes: removing the first proportion of the sequence numbers in the importance ranking sequence number to obtain the less important sub-sequence, in response to the importance ranking sequence number being in ascending order of importance.

[0069] Optionally, the method for determining the less important sub-sequence based on the importance ranking sequence number and the proportion further includes: removing the last proportion of the sequence numbers in the importance ranking sequence number to obtain the less important sub-sequence, in response to the importance ranking sequence number being in descending order of importance.

[0070] The method for determining the less important sub-sequence provided in the optional implementation determines the proportion of the non-important coded symbols based on the diversity ratio, and determines the less important sub-sequence based on the importance ranking sequence number and the proportion, thereby providing a reliable implementation for obtaining the less important sub-sequence.

[0071] Figure 4 A flow 400 of another embodiment of the semantic communication method according to the present disclosure is shown, and the semantic communication method includes the following steps:

[0072] In step 401, a target coded symbol sequence transmitted through different independent fading paths is received.

[0073] In the embodiment, the independent fading path is a fading channel for independently transmitting a signal. In the process of transmitting the signal, the target coded symbol sequence transmitted through different independent fading paths has multiple fading paths. The target coded symbols with the same semantics in the target coded symbol sequence can be transmitted in different fading channels. Since the same signal is lost differently in different fading channels, the received multiple statistically independent fading signals in different fading channels can be combined (selected and combined) by the receiving end to reduce the influence of the fading channel.

[0074] In this embodiment, the execution subject on which the semantic communication method of the present disclosure operates can be a receiving end.

[0075] In step 402, the first semantic code symbol with the same semantics in the target code symbol sequence is merged.

[0076] In this embodiment, the semantic code symbols with the same semantics can be merged by using appropriate merging methods, and the merging methods can be selected according to the independent fading paths adopted. The merging of the semantic code symbols with the same semantics refers to the merging of the semantic code symbols representing the same content or the same semantics. The merging methods can be different based on the independent fading paths. The merging process is a traditional processing method, which will not be described here.

[0077] As shown in Figure 5 , the OFDM receiver in the receiving end receives the OFDM symbol Os sent by the sending end, performs OFDM demapping on the OFDM symbol Os, obtains the target code symbol sequence Sd transmitted through different independent fading paths, and obtains the receiving code symbol sequence M after processing the target code symbol sequence Sd.

[0078] In step 403, the intermediate code symbol sequence is obtained based on the first semantic code symbol and the second semantic code symbol without the same semantics in the target code symbol sequence.

[0079] As shown in Figure 5 , all the semantic code symbols in the target code symbol sequence Sd enclosed in the braces are the first semantic code symbols. The target code symbol sequence Sd has two identical first semantic code symbols. The code symbol sequence in the target code symbol sequence Sd that is directly transitioned to the receiving code symbol sequence M through the straight arrow is the second semantic code symbol. The step 403 includes: combining the first semantic code symbol and the second semantic code symbol according to the pre-set semantic code symbol order to obtain the intermediate code symbol sequence.

[0080] In step 404, the target data corresponding to the source data is obtained based on the intermediate code symbol sequence and the first semantic code symbol.

[0081] In this embodiment, the target data corresponding to the source data is the signal actually received in the receiving end. Therefore, the target data corresponding to the source data can be the source data or data equivalent to the source data. As shown in Figure 5 , the target data is the data obtained after the decoder D in the receiving end decodes the receiving code symbol sequence M, wherein the receiving code symbol sequence M is the sequence corresponding to the semantic code symbol sequence obtained after the target code symbol sequence Sd is encoded and processed.

[0082] In the embodiment, the step 403 comprises: decoding the intermediate coded symbol sequence to obtain intermediate data, adding zero values with the same number of bits as the first semantic coded symbol to the intermediate data to obtain target data corresponding to the source data.

[0083] Optionally, the step 403 further comprises: performing zero padding processing on the intermediate coded symbol sequence by the same number of bits as the first semantic coded symbol to obtain a processed coded symbol sequence; and performing decoding processing on the processed coded symbol sequence to obtain target data corresponding to the source data.

[0084] The semantic communication method provided in the embodiment receives a target coded symbol sequence transmitted through different independent fading paths, merges first semantic coded symbols with the same semantics in the target coded symbol sequence, obtains a target coded symbol sequence based on the first semantic coded symbols and second semantic coded symbols without the same semantics in the target coded symbol sequence, and obtains target data corresponding to the source data based on the intermediate coded symbol sequence and the first semantic coded symbol, thereby providing a reliable implementation manner for obtaining the target data.

[0085] In some optional implementations of the disclosure, the intermediate data corresponding to the source data is obtained based on the intermediate coded symbol sequence and the first semantic coded symbol, which comprises: determining unimportant coded symbols based on the first semantic coded symbol; splicing the unimportant coded symbols and the intermediate coded symbol sequence to obtain a received coded symbol sequence; and obtaining target data corresponding to the source data based on the received coded symbol sequence.

[0086] In the optional implementation, the unimportant coded symbol is a symbol that needs to be padded to the intermediate coded symbol sequence, and the value obtained after decoding the symbol can help obtain the target data. The unimportant coded symbol can be directly set as zero with a certain number of bits, and the received coded symbol sequence can be obtained by directly padding the unimportant coded symbol to the intermediate coded symbol sequence.

[0087] In the optional implementation, the received coded symbol sequence is a symbol sequence obtained after splicing the unimportant coded symbols and the intermediate coded symbol sequence. The splicing position of the unimportant coded symbols in the intermediate coded symbol sequence can be determined based on the importance ordering sequence of the semantic coded symbols in the intermediate coded symbol sequence. When the importance ordering sequence of the semantic coded symbols in the intermediate coded symbol sequence is in descending order, the unimportant coded symbols are spliced at the rear of the intermediate coded symbol sequence to form the received coded symbol sequence. When the importance ordering sequence of the semantic coded symbols in the intermediate coded symbol sequence is in ascending order, the unimportant coded symbols are spliced at the front of the intermediate coded symbol sequence to form the received coded symbol sequence.

[0088] Optionally, the determining the non-important encoding symbols based on the first semantic encoding symbols comprises: generating the non-important encoding symbols based on a distribution of the first semantic encoding symbols.

[0089] Optionally, the determining the non-important encoding symbols based on the first semantic encoding symbols comprises: inputting the first semantic encoding symbols into a generative AI model (such as a GAN model, a VAE model, and a Diffusion model) to obtain the non-important encoding symbols generated by the generative AI model.

[0090] In the optional implementation, the obtaining the target data corresponding to the source data based on the received encoding symbol sequence comprises: performing decoding processing on the received encoding symbol sequence to obtain the target data corresponding to the source data. The decoding processing is a process of restoring a number into the content it represents or converting an electrical pulse signal, an optical signal, a radio wave, etc. into the information, data, etc. it represents. Decoding is a process of restoring the received symbols or codes by the recipient into information, which corresponds to the encoding process of the target encoding symbol sequence.

[0091] As one of the practical applications of the present disclosure, Figure 3 、 Figure 5 The semantic communication method is demonstrated in the semantic communication process of Massive MIMO CSI feedback. In this process, the CSI of the downlink is the source data. The specific steps are as follows:

[0092] 1) The CSI data of the downlink as the source data, first obtains the corresponding semantic encoding symbol sequence S through the semantic encoder E. The semantic encoder E uses a mask-based adaptive deep coding training framework for end-to-end training, so that the semantic encoder E actively learns and naturally performs importance sorting according to the spatial position of the semantic symbol, to obtain the importance sorting sequence number.

[0093] 2) According to the importance sorting sequence number, the non-important encoding symbols at the back of the sorting are selectively discarded, and the OFDM subcarriers vacated will be redistributed for frequency diversity transmission of the important semantic symbols at the front of the sorting. This means that the same important semantic symbols will be allocated to different OFDM subcarriers for transmission to enhance their reliability.

[0094] 3) The target encoding symbol sequence is sent to the receiving end. In this process, the important semantic symbols will be frequency diversity transmitted, and multiple subcarriers will be used to increase the robustness of the transmission.

[0095] 4) The receiving end receives the semantic encoding symbols on different frequency subcarriers through the wireless channel. The receiving end combines the same semantic semantic encoding symbols diversity transmitted on different frequency subcarriers to obtain the target encoding symbol sequence.

[0096] 5) For the previously discarded non-important encoding symbols, the receiving end generates by using the zero padding method and splices with the merged important semantic symbols to form a complete receiving encoding symbol sequence M.

[0097] 6) The receiving end recovers the receiving encoding symbol sequence M into the target data corresponding to the original CSI data based on the semantic decoder D. In this way, the receiving end can obtain more reliable CSI information than the traditional method, and improve the overall recovery performance of the communication system.

[0098] Through the semantic importance-based diversity method, in the semantic communication process of the Massive MIMO CSI feedback, the important semantic symbols are transmitted by frequency diversity in the OFDM subcarriers, and the data received by different frequency subcarriers at the receiving end is merged, which effectively improves the recovery degree of the receiving end, especially in the case of poor channel conditions, which can better reflect its advantages.

[0099] Further reference Figure 6 , as the implementation of the method shown in the above figures, the disclosure provides an embodiment of a semantic communication device, which corresponds to the method embodiment shown in Figure 1 , and the device can be specifically applied to various electronic devices.

[0100] As shown in Figure 6 , the semantic communication device 600 provided in the embodiment includes an acquisition unit 601, a determination unit 602, a obtaining unit 603, and a transmission unit 604. The acquisition unit 601 can be configured to acquire source data. The determination unit 602 can be configured to determine a semantic encoding symbol sequence and an importance ranking sequence number of a semantic encoding symbol in the semantic encoding symbol sequence based on the source data. The obtaining unit 603 can be configured to obtain a target encoding symbol sequence based on the importance ranking sequence number and the semantic encoding symbol sequence. The transmission unit 604 can be configured to transmit the target encoding symbol sequence to a receiving end through different independent fading paths, so that the receiving end obtains target data corresponding to the source data.

[0101] In the embodiment, the specific processing of the acquisition unit 601, the determination unit 602, the obtaining unit 603, and the transmission unit 604 in the semantic communication device 600 and the technical effects brought by the specific processing can be respectively referred to the related description of the steps 101, 102, 103, and 104 in the corresponding embodiment. Figure 1 The corresponding embodiment in the corresponding embodiment, the steps 101, 102, 103, and 104 are not described here.

[0102] In some optional implementations of the present embodiment, the determination unit 602 is configured to determine the data type based on the source data, determine the semantic space position model based on the data type, and input the source data into the semantic space position model to obtain the semantic encoding symbol sequence output by the semantic space position model and the importance ranking sequence number of the semantic encoding symbols in the semantic encoding symbol sequence.

[0103] In some optional implementations of the present embodiment, the determination unit 602 is configured to determine the data type based on the source data, determine the semantic encoder based on the data type, input the source data into the semantic encoder to obtain the semantic output of the semantic encoder, and input the semantic encoding symbol sequence into the pre-set importance ranking model to obtain the importance ranking sequence number of the semantic encoding symbols in the semantic encoding symbol sequence output by the importance ranking model.

[0104] In some optional implementations of the present embodiment, the obtaining unit 603 is configured to determine the less important sub-sequence number based on the importance ranking sequence number, remove the semantic encoding symbols corresponding to the less important sub-sequence number in the semantic encoding symbol sequence to obtain an intermediate encoding symbol sequence, determine the most important sub-sequence number having the same number of bits as the less important sub-sequence number based on the importance ranking sequence number, and supplement the semantic encoding symbols corresponding to the most important sub-sequence number into the intermediate encoding symbol sequence to obtain the target encoding symbol sequence.

[0105] In some optional implementations of the present embodiment, the obtaining unit 603 is further configured to determine the proportion of the unimportant encoding symbols based on the diversity ratio, and determine the less important sub-sequence number based on the importance ranking sequence number and the proportion.

[0106] The semantic communication method provided by the embodiments of the present disclosure first acquires source data, then determines a semantic encoding symbol sequence and an importance ranking sequence number of the semantic encoding symbols in the semantic encoding symbol sequence based on the source data, further obtains a target encoding symbol sequence based on the importance ranking sequence number and the semantic encoding symbol sequence, and finally transmits the target encoding symbol sequence to a receiving end through different independent fading paths, so that the receiving end obtains target data corresponding to the source data. In this way, by reordering and resource allocation of the semantic encoding symbols in combination with the importance characteristics of the semantic encoding symbols, the target encoding symbol sequence is determined, and the target encoding symbol sequence is transmitted in diversity, thereby preferentially guaranteeing the reliable transmission of key information, improving the recovery performance of information, and being particularly suitable for use in an environment with poor channel conditions, and significantly improving the reliability and accuracy of communication.

[0107] In the technical solutions of the present disclosure, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information comply with relevant laws and regulations and do not violate public order and good customs.

[0108] Further referring to Figure 7 , as an implementation of the method shown in the above figures, the disclosure provides another embodiment of a semantic communication device, which device embodiment corresponds to the method embodiment shown in Figure 4 , and the device can be specifically applied in various electronic devices.

[0109] As shown in Figure 7 , the semantic communication device 700 provided in the embodiment includes a receiving unit 701, a merging unit 702, a combining unit 703, and a restoring unit 704. The receiving unit 701 can be configured to receive a target coded symbol sequence transmitted through different independent fading paths. The merging unit 702 can be configured to merge first semantic coded symbols with the same semantics in the target coded symbol sequence. The combining unit 703 can be configured to obtain a target coded symbol sequence based on the first semantic coded symbols and second semantic coded symbols in the target coded symbol sequence that do not have the same semantics. The restoring unit 704 can be configured to obtain target data corresponding to the source data based on the intermediate coded symbol sequence and the first semantic coded symbols.

[0110] In the embodiment, the specific processing of the receiving unit 701, the merging unit 702, the combining unit 703, and the restoring unit 704 in the semantic communication device 700 and the technical effects brought by the specific processing can be respectively referred to the related descriptions of the steps 401, the step 402, the step 403, and the step 404 in the corresponding embodiment, which will not be described here. Figure 4 The corresponding embodiment corresponds to the method embodiment shown in

[0111] In some optional implementations of the embodiment, the restoring unit 703 is configured to determine unimportant coded symbols based on the first semantic coded symbols, splice the unimportant coded symbols with the intermediate coded symbol sequence to obtain a received coded symbol sequence, and obtain target data corresponding to the source data based on the received coded symbol sequence.

[0112] Further referring to Figure 8 , as an implementation of the method shown in the above figures, the disclosure provides another embodiment of a semantic communication system, which system embodiment corresponds to the method embodiment shown in Figure 1 , Figure 4 , and the system can be specifically applied in various electronic devices.

[0113] As shown in Figure 8As shown, the semantic communication system 800 provided by the embodiment includes a sending end 801 and a receiving end 802. The sending end 801 is configured to acquire source data, determine a semantic code symbol sequence and an importance order sequence of semantic code symbols in the semantic code symbol sequence based on the source data, obtain a target code symbol sequence based on the importance order sequence and the semantic code symbol sequence, and transmit the target code symbol sequence to the receiving end through different independent fading paths. The receiving end 802 is configured to receive the target code symbol sequence transmitted through the different independent fading paths, combine first semantic code symbols with the same semantics in the target code symbol sequence, obtain a target code symbol sequence based on the first semantic code symbols and second semantic code symbols without the same semantics in the target code symbol sequence, and obtain target data corresponding to the source data based on the intermediate code symbol sequence and the first semantic code symbols.

[0114] In the embodiment, the specific process of the sending end 801 in the semantic communication system 800 and the technical effects brought by the specific process can be referred to the specific process of the sending end 801 and the technical effects brought by the specific process of the corresponding embodiment respectively. Figure 1 The specific process of the receiving end 802 in the embodiment and the technical effects brought by the specific process can be referred to the specific process of the receiving end 802 and the technical effects brought by the specific process of the corresponding embodiment respectively. Figure 4 The related description of the steps 401, 402 and 403 in the corresponding embodiment is not described here again.

[0115] According to the embodiments of the present disclosure, the present disclosure further provides an electronic device, a readable storage medium and a computer program product.

[0116] Figure 9 A schematic block diagram of an example electronic device 900 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present disclosure described and / or claimed in this document.

[0117] As Figure 9As shown, the device 900 includes a computing unit 901 that can perform various appropriate actions and processes in accordance with a computer program stored in a read-only memory (ROM) 902 or a computer program loaded from a storage unit 908 into a random access memory (RAM) 903. In the RAM 903, various programs and data required for the operation of the device 900 can also be stored. The computing unit 901, the ROM 902, and the RAM 903 are connected to each other through a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0118] A plurality of components in the device 900 are connected to the I / O interface 905, including: an input unit 906, such as a keyboard, a mouse, etc.; an output unit 907, such as various types of displays, speakers, etc.; a storage unit 908, such as a magnetic disk, an optical disk, etc.; and a communication unit 909, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 909 allows the device 900 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks.

[0119] The computing unit 901 can be various general and / or special purpose processing components with processing and computing capabilities. Some examples of the computing unit 901 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 901 performs various methods and processes described above, such as the semantic communication method. For example, in some embodiments, the semantic communication method can be implemented as a computer software program that is tangibly embodied in a machine-readable medium, such as the storage unit 908. In some embodiments, part or all of the computer program can be loaded and / or installed on the device 900 via the ROM 902 and / or the communication unit 909. When the computer program is loaded into the RAM 903 and executed by the computing unit 901, one or more steps of the semantic communication method described above can be performed. Alternatively, in other embodiments, the computing unit 901 can be configured to perform the semantic communication method by any other appropriate means, such as by means of firmware.

[0120] The various embodiments of the systems and techniques described above can be implemented in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a load programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0121] Program code for carrying out methods of the present disclosure can be written in any combination of one or more programming languages. The program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, implements the functions / acts specified in the flowcharts and / or block diagrams. The program code can be executed entirely on a machine, partially on a machine, partially on a machine as a stand-alone software package, or entirely on a remote machine or server.

[0122] In the context of the present disclosure, a machine-readable medium can be a tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include but is not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0123] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0124] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a client computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0125] The computer system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server is generally established by computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0126] It should be understood that the various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the scope of the present disclosure. For example, the steps recited in the present disclosure can be performed in parallel, in series, or in a different order, without departing from the desired results of the technology disclosed in the present disclosure, and are not limited herein.

[0127] The specific embodiments described above are not intended to be limiting, and persons skilled in the art will appreciate that various modifications, combinations, sub-combinations and alternatives can be made to the specific embodiments without departing from the spirit and principles of the disclosure. Accordingly, the disclosure is not limited to the specific embodiments described above.

Claims

1. A semantic communication method, the method comprising: obtaining source data; determining a sequence of semantic code symbols and an importance order sequence of the semantic code symbols in the sequence of semantic code symbols based on the source data; obtaining a sequence of target code symbols based on the importance order sequence and the sequence of semantic code symbols; transmitting the sequence of target code symbols to a receiving end through different independent fading paths, so that the receiving end obtains target data corresponding to the source data; the obtaining of the sequence of target code symbols based on the importance order sequence and the sequence of semantic code symbols comprises: determining a sub-sequence of less important code symbols based on the importance order sequence; obtaining a sequence of intermediate code symbols by removing the semantic code symbols corresponding to the sub-sequence of less important code symbols from the sequence of semantic code symbols; determining a sub-sequence of most important code symbols having the same number of bits as the sub-sequence of less important code symbols based on the importance order sequence; obtaining the sequence of target code symbols by supplementing the semantic code symbols corresponding to the sub-sequence of most important code symbols after the sequence of intermediate code symbols.

2. The method of claim 1, wherein, the determining of the sequence of semantic code symbols and the importance order sequence of the semantic code symbols in the sequence of semantic code symbols based on the source data comprises: determining a data type based on the source data; determining a semantic space position model based on the data type; inputting the source data into the semantic space position model to obtain a sequence of semantic code symbols output by the semantic space position model and an importance order sequence of the semantic code symbols in the sequence of semantic code symbols.

3. The method of claim 1, wherein, the determining of the sequence of semantic code symbols and the importance order sequence of the semantic code symbols in the sequence of semantic code symbols based on the source data comprises: determining a data type based on the source data; determining a semantic encoder based on the data type; inputting the source data into the semantic encoder to obtain a sequence of semantic code symbols output by the semantic encoder; inputting the sequence of semantic code symbols into a pre-set importance order model to obtain an importance order sequence of the semantic code symbols in the sequence of semantic code symbols output by the importance order model.

4. The method of claim 1, wherein, the determining of the sub-sequence of less important code symbols based on the importance order sequence comprises: determining a proportion of non-important code symbols based on a diversity ratio; determining the sub-sequence of less important code symbols based on the importance order sequence and the proportion. 5.A semantic communication method, the method comprising: receiving a sequence of target code symbols transmitted through different independent fading paths, the sequence of target code symbols being obtained based on the semantic communication method of any one of claims 1-4; merging first semantic code symbols having the same semantics in the sequence of target code symbols; obtaining a sequence of intermediate code symbols based on the first semantic code symbols and second semantic code symbols not having the same semantics in the sequence of target code symbols; obtaining target data corresponding to source data based on the sequence of intermediate code symbols and the first semantic code symbols.

6. The method of claim 5, wherein, The obtaining, based on the intermediate code symbol sequence and the first semantic code symbol, of target data corresponding to the source data comprises: determining, based on the first semantic code symbol, non-important code symbols; splicing the non-important code symbols and the intermediate code symbol sequence to obtain a receiving code symbol sequence; obtaining, based on the receiving code symbol sequence, target data corresponding to the source data.

7. A semantic communication apparatus, the apparatus comprising: an obtaining unit configured to obtain source data; a determining unit configured to determine, based on the source data, a semantic code symbol sequence and an importance order number of a semantic code symbol in the semantic code symbol sequence; an obtaining unit configured to determine, based on the importance order number, a sub-importance number; remove the semantic code symbol corresponding to the sub-importance number in the semantic code symbol sequence to obtain an intermediate code symbol sequence; determine, based on the importance order number, a most important sub-number having the same number of bits as the sub-importance number; and supplement the semantic code symbol corresponding to the most important sub-number after the intermediate code symbol sequence to obtain a target code symbol sequence; a transmitting unit configured to transmit the target code symbol sequence to a receiving end through different independent fading paths, so that the receiving end obtains target data corresponding to the source data.

8. A semantic communication apparatus, the apparatus comprising: a receiving unit configured to receive a target code symbol sequence transmitted through different independent fading paths, the target code symbol sequence being obtained based on the semantic communication method of any one of claims 1-4; a merging unit configured to merge first semantic code symbols having the same semantics in the target code symbol sequence; a combining unit configured to obtain, based on the first semantic code symbols and second semantic code symbols not having the same semantics in the target code symbol sequence, an intermediate code symbol sequence; a restoring unit configured to obtain, based on the intermediate code symbol sequence and the first semantic code symbol, target data corresponding to the source data.

9. A semantic communication system, the system comprising: a receiving end and a sending end; the sending end is configured to obtain source data; determine, based on the source data, a semantic code symbol sequence and an importance order number of a semantic code symbol in the semantic code symbol sequence; and determine, based on the importance order number, a sub-importance number; remove the semantic code symbol corresponding to the sub-importance number in the semantic code symbol sequence to obtain an intermediate code symbol sequence; determine, based on the importance order number, a most important sub-number having the same number of bits as the sub-importance number; supplement the semantic code symbol corresponding to the most important sub-number after the intermediate code symbol sequence to obtain a target code symbol sequence; and transmit the target code symbol sequence to a receiving end through different independent fading paths. The receiving end is configured to receive target code symbol sequences transmitted through different independent fading paths, combine first semantic code symbols with the same semantics in the target code symbol sequences, obtain an intermediate code symbol sequence based on the first semantic code symbols and second semantic code symbols without the same semantics in the target code symbol sequences, and obtain target data corresponding to source data based on the intermediate code symbol sequence and the first semantic code symbols.

10. An electronic device, comprising: The method comprises: at least one processor; and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform any one of claims 1-6.

11. A non-transitory computer-readable storage medium having stored thereon computer instructions, wherein, The computer instructions are configured to enable the computer to perform any one of claims 1-6.

Images