Semantic communication method, system and device and storage medium

Abstract

The invention provides a semantic communication method, system and device and a storage medium, and relates to the technical field of communication. According to the specific implementation scheme, a sending end carries out encryption semantic coding processing and weight configuration on information source information, and matched block encryption semantic vectors and importance indication information are obtained; performing weighted coding processing on the matched block encryption semantic vector and the importance indication information to obtain block encryption semantic information; based on information verification coding processing, the grouping encryption semantic information is converted into grouping verification encryption semantic information, the grouping verification encryption semantic information is transmitted to a receiving end, and the receiving end carries out corresponding decryption and decoding on data to obtain information source information. The security, the reliability and the transmission efficiency of information transmission are ensured at the same time, the communication system structure is optimized, and the communication cost is saved at the same time.

Description

Technical Field

[0001] The present disclosure relates to the field of communication technology, and in particular to a semantic communication method, system, device and storage medium. Background Art

[0002] Semantic communication is an intelligent communication method that significantly improves communication system performance by deeply understanding the content and underlying intent of information, enabling efficient and accurate data transmission and interaction. It will play a vital role in the 6G era. Introducing semantic communication technology into communication systems requires attention to the security, reliability, and efficiency of information transmission. However, encryption can destroy the correlation between information, resulting in distortion after decoding. The combination of existing mainstream encoding methods and encryption technologies increases the computational burden on both the sender and receiver, increasing the complexity of the communication system and the cost of cross-layer information transmission, while reducing the performance and transmission efficiency of the communication system. Summary of the Invention

[0003] The present disclosure provides a semantic communication method, system, device and storage medium.

[0004] According to a first aspect of the present disclosure, a semantic communication method is provided, which is applied to a sending end and includes:

[0005] Performing encrypted semantic coding and weight configuration on the source information to obtain matching block encrypted semantic vectors and importance indication information;

[0006] performing weighted coding processing on the matched block encryption semantic vector and the importance indication information to obtain block encryption semantic information;

[0007] Based on information verification coding processing, the block encrypted semantic information is converted into block verification encrypted semantic information;

[0008] The packet verification encryption semantic information is transmitted to a receiving end.

[0009] According to a second aspect of the present disclosure, a semantic communication device is provided, which is applied to a sending end and includes:

[0010] The semantic coding module is used to perform encrypted semantic coding processing and weight configuration on the source information to obtain the matching group encryption semantic vector and importance indication information;

[0011] A weighted coding module, configured to perform weighted coding processing on the matched block encryption semantic vector and the importance indication information to obtain block encryption semantic information;

[0012] A checksum coding module, configured to convert the block encrypted semantic information into block checksum encrypted semantic information based on information checksum coding processing;

[0013] The sending module is used to transmit the group verification encryption semantic information to the receiving end.

[0014] According to a third aspect of the present disclosure, a semantic communication method is provided, which is applied to a receiving end and includes:

[0015] Receive the packet verification encryption semantic information sent by the sender;

[0016] Based on information verification decoding processing, the group verification encryption semantic information is converted into group encryption semantic information;

[0017] Decapsulating the block encrypted semantic information to obtain a block encrypted semantic vector;

[0018] Perform semantic decryption and decoding processing on the block encrypted semantic vector to obtain source information.

[0019] According to a fourth aspect of the present disclosure, a semantic communication device is provided, applied to a receiving end, comprising:

[0020] A receiving module, configured to receive the packet verification encrypted semantic information sent by the sending end;

[0021] A checksum decoding module, configured to convert the group checksum encrypted semantic information into group encrypted semantic information based on information checksum decoding processing;

[0022] a decapsulation module, configured to decapsulate the block encrypted semantic information to obtain a block encrypted semantic vector;

[0023] The semantic decoding module is used to perform semantic decryption decoding processing on the block encrypted semantic vector to obtain source information.

[0024] According to the fifth aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; 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 execute the semantic communication method described in the first aspect of the present disclosure, or to execute the semantic communication method described in the third aspect of the present disclosure.

[0025] According to the sixth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the semantic communication method described in the first aspect of the present disclosure, or to execute the semantic communication method described in the third aspect of the present disclosure.

[0026] According to the seventh aspect of the present disclosure, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the semantic communication method according to the first aspect of the present disclosure, or executes the semantic communication method according to the third aspect of the present disclosure.

[0027] The technology disclosed in the present invention solves the problems in existing semantic communication, such as the difficulty in information encryption, distortion after information decryption, low transmission efficiency, and high communication costs. At the same time, it ensures the security, reliability and transmission efficiency of information transmission, optimizes the communication system structure and saves communication costs.

[0028] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The accompanying drawings are provided to facilitate a better understanding of the present invention and do not constitute a limitation of the present disclosure.

[0030] Figure 1 is an interactive schematic diagram of the semantic communication method provided by the present disclosure;

[0031] Figure 2 is a cross-layer design flow chart of the semantic communication method provided by the present disclosure;

[0032] Figure 3 is a flowchart of a semantic communication method provided by the first embodiment of the present disclosure;

[0033] Figure 4 is a flowchart of information processing at the sending end provided by the first embodiment of the present disclosure;

[0034] Figure 5 This is a flow chart of importance indication information conversion provided by the first embodiment of the present disclosure;

[0035] Figure 6 This is a schematic diagram of a training method for common quantized source coding provided by the first embodiment of the present disclosure;

[0036] Figure 7 This is a schematic diagram of the application flow of a common quantized signal source codec at both receiving and receiving ends provided by the first embodiment of the present disclosure;

[0037] Figure 8 This is a schematic diagram of a training method for vector / tensor quantization source coding provided by the first embodiment of the present disclosure;

[0038] Figure 9 This is a schematic diagram of the application flow of the vector / tensor quantization source codec for the receiving parties provided in the first embodiment of the present disclosure.

[0039] Figure 10 is a structural block diagram of a semantic communication device provided by the second embodiment of the present disclosure;

[0040] Figure 11 is a flowchart of a semantic communication method provided by the third embodiment of the present disclosure;

[0041] Figure 12 is a flowchart of information processing at the receiving end provided by the third embodiment of the present disclosure;

[0042] Figure 13 is a structural block diagram of a semantic communication device provided by the fourth embodiment of the present disclosure;

[0043] Figure 14 It is a block diagram of an electronic device provided in the fifth embodiment of the present disclosure and used to implement one of the semantic communication methods in the embodiments of the present disclosure. DETAILED DESCRIPTION

[0044] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0045] First, in semantic communication, joint source-channel coding is the mainstream coding method. However, this approach presents encryption challenges. For example, encrypting information before encoding can destroy semantic relevance, while encrypting information after encoding is affected by channel quality, potentially leading to decryption errors and reduced transmission reliability. Furthermore, the introduction of encryption significantly increases the computational burden on both the sender and receiver, impacting system performance and transmission efficiency. However, neglecting to encrypt transmitted information creates the risk of eavesdropping and tampering, making it impossible to ensure secure transmission.

[0046] Secondly, in order to reduce the computational burden and transmission cost of the communication system and improve transmission efficiency, it is necessary to fully consider the importance of the information being transmitted to avoid the unnecessary waste of communication system resources caused by indiscriminately triggering the retransmission mechanism when information transmission errors occur, resulting in reduced transmission efficiency and increased communication costs.

[0047] Thirdly, joint source-channel encoders generally require the transmission of channel state information at the physical layer, which significantly increases the complexity of the communication system and the cost of cross-layer information transmission. Therefore, it is necessary to consider the end-to-end transmission process across the application layer, network layer, data link layer, and physical layer in actual communication systems.

[0048] In order to ensure the security, reliability and transmission efficiency of information transmission and reduce communication costs in semantic communication, the present disclosure provides a semantic communication method, system, device and storage medium based on the above-mentioned comprehensive consideration of encryption difficulty, deployment cost and system complexity.

[0049] Figure 1 This is an interactive diagram of the semantic communication method provided by the present disclosure. Figure 1 The overall process of the semantic communication method provided by the present disclosure is as follows:

[0050] The sending end performs encryption semantic coding processing and weight configuration on the source information to obtain a matching group encryption semantic vector and importance indication information; performs weighted coding processing on the matching group encryption semantic vector and the importance indication information to obtain group encryption semantic information; based on information verification coding processing, the group encryption semantic information is converted into group verification encryption semantic information; and the group verification encryption semantic information is transmitted to the receiving end through a physical transmission channel.

[0051] The receiving end receives the packet verification encrypted semantic information sent by the sending end; based on the information verification decoding processing, the packet verification encrypted semantic information is converted into packet encryption semantic information; the packet encryption semantic information is decapsulated to obtain a packet encryption semantic vector; and the packet encryption semantic vector is semantically decrypted and decoded to obtain source information.

[0052] The present disclosure achieves efficient data compression by extracting and restoring the semantics of source information, preserving semantic relevance, and combating quantization noise. Efficient data compression ensures accurate transmission of information semantics while minimizing the use of system resources, thereby improving transmission efficiency and reducing transmission costs. By quantizing and encoding source information, semantic communication can be integrated with conventional communication systems, enabling conventional encryption, channel coding, and modulation methods to work in conjunction with the source codec, improving the performance and applicability of the overall communication system. The encryption and decryption of semantic information ensures the security of information transmission. By hierarchically extracting and reorganizing source information, accurate information capture and effective integration are achieved. Errors are promptly identified during information verification, and based on the precise assessment of the importance of information at different levels and dimensions, the retransmission of errors can be flexibly selected. This ensures the reliability of critical information more intelligently during transmission, achieving more efficient and accurate information transmission. The steps provided in the present disclosure can be performed at different transmission layers on the transmitting and receiving ends of a communication system. By effectively allocating tasks to different transmission layers of the communication system, the modular design of the system is optimized, thereby improving the performance of the communication system.

[0053] The cross-layer design solution provided by this disclosure is as follows: Figure 2 shown.

[0054] The semantic communication method, apparatus, device, and storage medium provided by the present disclosure are described below with reference to the accompanying drawings.

[0055] In the technical solutions disclosed herein, the acquisition, storage, and application of user personal information involved comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0056] First embodiment

[0057] Figure 3 One of the flowcharts of the semantic communication method provided in the embodiment of the present disclosure; Figure 4 This is a flowchart of information processing at the sending end provided in an embodiment of the present disclosure.

[0058] Reference Figure 3 and Figure 4 The present disclosure provides a semantic communication method, which is applied to a sending end. The method may include steps 101 to 104.

[0059] Step 101 : Perform encryption semantic coding and weight configuration on the source information to obtain a matching group encryption semantic vector and importance indication information.

[0060] In this step, by extracting source semantic information, each encrypted semantic vector is obtained, preserving semantic relevance and combating quantization noise, achieving efficient data compression. This efficient data compression ensures accurate semantic transmission while minimizing the use of system resources, thereby improving transmission efficiency and reducing transmission costs. Weighted configuration generates importance indicators that are matched to each encrypted semantic vector. In subsequent steps, this importance indicator is used to identify the importance of erroneous information and flexibly select whether to retransmit errors, thereby improving information transmission reliability and efficiency.

[0061] Specifically, this step can be applied to the fourth layer of the sending end, that is, the application layer of the sending end, and includes steps 201 to 205.

[0062] Step 201: extracting semantic features of the source information.

[0063] In this step, the core meaning of the information is captured through semantic extraction, and the information to be transmitted is converted into semantic features so that the information content can be transmitted and understood efficiently and accurately.

[0064] Step 202 : Based on a preset weight allocation model, weights are assigned to the semantic features according to the importance levels of the semantic features to obtain the importance indication information.

[0065] In this step, based on the preset weight distribution model, corresponding weights are configured for semantic features according to their importance levels, thereby accurately reflecting the importance of each feature in the information and providing importance indications for subsequent error retransmissions.

[0066] Optionally, the preset weight distribution model includes but is not limited to: an information entropy importance distribution model and / or an importance experience distribution model.

[0067] The information entropy importance allocation model can more accurately and objectively determine the importance level of information, and the importance experience allocation model is more conducive to time cost and computing cost. The choice of model is determined according to the needs of the actual application scenario and is not limited here.

[0068] Specifically, different semantic features are assigned different importance levels. This importance level can be simply assigned based on importance experience, or calculated using an entropy model or other similar models. When using a model to calculate the importance level based on the correlation between different semantic information, refer to the attached Figure 5 , the specific calculation method is as follows:

[0069] First, the information entropy calculation model is used to obtain the information entropy of the semantic feature based on the correlation between semantic information. Then the information entropy is normalized, and the maximum information entropy is set as max, the minimum information entropy is set as min, and the value of each information entropy is e ij . It can be normalized by the following formula:

[0070] Then, the information entropy of the same feature is averaged to obtain the group normalized information entropy. Finally, the information entropy is divided into different value ranges between [0, 1] according to the pre-set importance feature mapping set, and different importance levels are assigned.

[0071] For example:

[0072] When the data value of the group normalized information entropy belongs to [0, 0.2), the importance indication information is set to level 1 (specific different levels may be represented by different data).

[0073] When the data value of the group normalized information entropy belongs to [0.2, 0.4), the importance indication information is set to level 2.

[0074] Similarly, when the data value of the group normalized information entropy belongs to [0.8, 1], the importance indication information is set to level 5.

[0075] Step 203: perform source coding and quantization processing on the semantic features to obtain semantic coding information.

[0076] By quantizing and encoding the source information, semantic communication can be combined with conventional communication systems, and conventional encryption, channel coding and modulation methods can work together with the source codec to improve the performance and applicability of the overall communication system.

[0077] Specifically, semantic features are source encoded and quantized. Since quantization will lose some precision, it should be taken into account when training the source codec to ensure the restoration effect.

[0078] Two ideas are provided here, corresponding to the quantization of real numbers and the quantization of vectors / tensors respectively.

[0079] Please refer to the attached Figure 6 , suppose the source information is encoded as after the encoder, and quantized as z after the quantizer q . Then the quantization noise is sg(z q -z), where sg(·) is used to remove the gradient information during training to avoid affecting the encoder training. The encoder / decoder trained after adding quantization noise can effectively combat the error caused by quantization. In the application process, the source encoding and decoding and quantization and dequantization process are as follows: Figure 7 shown.

[0080] The model construction during the training process is as follows Figure 8 As shown, a set of trainable codebooks is pre-set, and the codebook contains m possible sub-codes: Suppose the source information is divided into n sub-codes after the encoder: z1, z2, ..., z n For the i-th subcode z i , the closest codebook is Then the quantization noise is The role of sg(·) is to remove the gradient information during training to avoid affecting the training of the encoder. Let the source information be x and the recovery information be In order to train the codebook, its loss function should be the sum of two parts: λ is a hyperparameter that can be adjusted between (0,1], and C(·) is the closest subcode in the query codebook. During the application process, the quantizer is used for each subcode z i Find the codebook that is closest to it in the codebook And encode it as j. The dequantizer then searches the jth codebook in the codebook And place it in the corresponding position. The source encoding and decoding and quantization and dequantization process is as follows Figure 9 shown.

[0081] Step 204: perform grouping and vectorization processing on the semantic coding information to obtain a grouped semantic vector.

[0082] The semantic coding information is grouped and quantified, and each group of data is independent of each other, which is conducive to the extraction and integration of information in subsequent steps.

[0083] Optionally, the grouping processing of the semantic coding information includes: setting grouping granularity according to an application scenario.

[0084] Setting fine-grained grouping according to application scenarios can better meet the needs of different scenarios and improve the targetedness of information processing and transmission efficiency.

[0085] For example, according to the division of scenarios, the semantic communication enhancement scenario can be divided into a higher granularity, and the semantic communication enablement scenario can be divided into a lower granularity.

[0086] Specifically, for example:

[0087] For three-dimensional image semantic coding data of C×H×W (where C is the number of data channels, H is the image height, and W is the image width), it can be divided into low-granularity of 2×2=4 areas or high-granularity of 4×4=16 areas in the H×W dimension.

[0088] For C×L two-dimensional text semantic coding data (where C is the number of data channels and L is the length of the coding sequence), it can be divided into low granularity with 100 coding data per group or high granularity with 10 coding data per group.

[0089] Specifically, for example:

[0090] For the groups obtained by dividing the image into regions with high granularity, the data in each group is still a The three-dimensional multi-channel data is The data is expanded in the dimension of , and then the two-dimensional data is expanded into a vector with a length of Semantic vector data.

[0091] For the groups obtained by high-grained text division, the data in each group is still a C×10 two-dimensional multi-channel data. The two-dimensional data is expanded into a vector to obtain semantic vector data with a length of 10C.

[0092] When semantically coded information is grouped, the obtained semantic grouping coded information may exist in the form of tensors or matrices. In order to simplify the processing and analysis of complex data, enable computers to understand and operate data more efficiently, and retain the key features of the data, it is necessary to convert tensors or matrices with multiple dimensions into vectors. Therefore, the semantic grouping coded information needs to be vectorized.

[0093] Optionally, vectorizing the semantic grouping coded information includes: first determining the dimensional direction of the tensor / matrix to be processed; then calculating the length of the new vector, which will be the size of the tensor along the selected dimension multiplied by the product of all other dimensions. For example, for a third-order tensor, if you choose to expand along the first dimension, then the length of the new vector will be the size of the first dimension multiplied by the product of the second and third dimensions. Then, starting from the first element of the selected dimension, the elements are placed into the new vector in the order of the dimension. Finally, verify whether the generated new vector correctly contains all the information of the original tensor.

[0094] Step 205: encrypt the group semantic vector to obtain the group encrypted semantic vector.

[0095] The grouped semantic vectors obtained above are encrypted using an encryption algorithm to obtain grouped encrypted semantic vectors. During encryption, the data of each group does not interfere with each other. The security of information transmission is guaranteed by encrypting and decrypting the semantic information.

[0096] Step 102: Perform weighted coding processing on the matched block encryption semantic vector and the importance indication information to obtain block encryption semantic information.

[0097] This step can be applied to the third layer of the sending end, namely the transport layer.

[0098] By performing weighted coding processing on the matching group encryption semantic vectors and the importance indication information, the generated group encryption semantic information can accurately evaluate the importance of information at different levels and dimensions, flexibly choose whether to retransmit the error code, and provide a reliable basis for the subsequent choice of whether to retransmit the error code.

[0099] Optionally, step 102 specifically includes: encoding the matched group encryption semantic vector, the importance indication information, and the packet header information to obtain the group encryption semantic information.

[0100] Specifically, first, the packet encryption semantic vector passed down from the application layer is split into vectors, and the header information is added in front of the split data. The header information includes the source address, destination address, etc.

[0101] Next, based on the above information, the importance indication information calculated and passed down by the application layer is added to its header to obtain the packet encryption semantic information 1...packet encryption semantic information n (n is a natural number), and different importance levels are assigned to each data packet.

[0102] The above-mentioned addition of importance indication information is carried out in groups. For example, after the group encryption semantic vector 1 is split to obtain multiple group encryption semantic vectors, the packet header is added. When adding the importance indication information, all the data obtained from the splitting of the group encryption semantic vector 1 will be added with the same importance indication information, and so on for the group encryption semantic vector 2.

[0103] Step 103: Based on information verification coding processing, the block encrypted semantic information is converted into block verification encrypted semantic information.

[0104] This step can be applied to the second layer of the sending end, namely the network layer.

[0105] Optionally, step 103 includes:

[0106] Based on the importance indication information, the group encryption semantic information, the cyclic redundancy check code and the importance information check code are information-encoded to obtain the group check encryption semantic information.

[0107] A cyclic redundancy check (CRC) is used to detect errors that may occur during data transmission or storage. By incorporating a CRC, it identifies received information as bit-error-free, improving data accuracy and reliability. By incorporating a significance check code, significance-indicating information is verified for transmission errors, preventing the receiver from misjudging the significance of distorted information and preventing the omission of important information.

[0108] Specifically, first, a CRC checksum is added to the end of the packet encrypted semantic data packet obtained above, based on the importance indication information in the header, to obtain a packet encrypted semantic frame. The manner in which the CRC checksum is added and the number of bits used depend on the importance indication information. If a packet has a high importance level, the number of bits in its CRC checksum will be longer or the number of CRC checks will be greater. If a packet has a low importance indication information, the number of bits in its CRC checksum will be shorter or the number of CRC checks will be less, thereby ensuring that the more important the data, the lower the probability of errors.

[0109] For example:

[0110] Example 1: Different CRC checksum bits are used based on the importance of the information. For example, CRC32 (i.e., 32-bit checksum) is used for information at level 5; CRC24 is used for information at level 4; CRC16 is used for information at level 3; CRC12 is used for information at level 2; and CRC8 is used for information at level 1.

[0111] Example 2: Different CRC check times are used based on the importance of the information. For example: 4 CRC checks are used for information of level 4; 3 CRC checks are used for information of level 3; 2 CRC checks are used for information of level 2; and 1 CRC check is used for information of level 1.

[0112] Example 3: Different check combinations are used based on the importance of the information. For example, more check methods (parity check, checksum) are used for information with higher importance.

[0113] At the same time, in order to ensure that the importance indication information will not be wrong, an important indication information check code will be added to the end of the block encryption semantic frame.

[0114] Step 104: transmit the packet verification encryption semantic information to the receiving end.

[0115] This step can be applied to the first layer of the transmitter, ie, the physical layer.

[0116] S104: Transmit the packet verification encryption semantic information to a receiving end.

[0117] Optionally, transmitting the packet check encryption semantic information to the receiving end includes: performing channel coding and signal modulation based on the packet check encryption semantic information to obtain packet encryption semantic modulation information; matching the transmission channel according to the importance indication information, and transmitting the packet encryption semantic modulation information to the receiving end.

[0118] According to the importance indication information, the transmission channel is matched, and channels with high signal-to-noise ratio, low bit error rate, large channel capacity and high signal quality index are allocated to signal transmission with high importance level, and signals with low importance level are allocated to channel transmission with corresponding quality level, which is beneficial to improving the reliability of information transmission and system resource utilization.

[0119] Specifically, first, the block encrypted semantic frame obtained above is channel-coded to obtain a block encrypted semantic channel code, and the channel code should have a certain error correction function.

[0120] Optionally, the channel coding includes but is not limited to at least one of the following error correction coding technologies: Hamming code, convolutional code, turbo code (Turbo code), polarization code. Hamming code can detect and correct multiple bit errors, but requires high computational efficiency. Convolutional code can have low coding complexity, but when the constraint length increases, the decoding complexity will increase exponentially by 2. Turbo code achieves error correction performance close to the Shannon limit, but the decoding complexity is high and requires a longer decoding delay. Polarization code: has strong error correction capability and noise resistance, but consumes more hardware resources. In actual applications, error correction coding is selected according to specific scenarios.

[0121] Next, the block-encrypted semantic channel code is modulated to generate block-encrypted semantic modulation information, which is then transmitted over the wireless channel. Different channel qualities are used for transmission based on the importance of each block-encrypted semantic modulation information. For example, for information of importance level 1, the channel with the best channel quality is selected for transmission. For information of importance level 2, the channel with the second-best channel quality is selected for transmission, and so on. Furthermore, the modulation scheme adapts to the current channel conditions to select the optimal modulation scheme.

[0122] The semantic communication method provided by the present disclosure ensures the security, reliability and efficiency of information transmission, optimizes the communication system structure and saves communication costs.

[0123] Second embodiment

[0124] Figure 10 This is one of the structural diagrams of the semantic communication device provided by the present disclosure. Figure 10 The semantic communication device 600 provided by the present disclosure is applied to a sending end and may include:

[0125] The semantic coding module 601 is used to perform encrypted semantic coding processing and weight configuration on the source information to obtain a matching block encrypted semantic vector and importance indication information.

[0126] The weighted coding module 602 is configured to perform weighted coding processing on the matched block encryption semantic vector and the importance indication information to obtain block encryption semantic information.

[0127] The check coding module 603 is configured to convert the block encrypted semantic information into block check encrypted semantic information based on information check coding processing.

[0128] The sending module 604 is configured to transmit the packet verification encryption semantic information to a receiving end.

[0129] The semantic communication device provided by the present disclosure ensures the security, reliability and efficiency of information transmission, optimizes the communication system structure and saves communication costs.

[0130] Third embodiment

[0131] Figure 11 This is one of the flow charts of the semantic communication method provided in the embodiment of the present disclosure, Figure 12 This is a flowchart of the information processing at the receiving end provided by the embodiment of the present disclosure. Figure 11 and Figure 12 The semantic communication method provided by the present disclosure is applied to the receiving end, and the method may include: steps 301 to 304.

[0132] Step 301: Receive packet verification encryption semantic information sent by a sending end.

[0133] This step can be applied to the first layer of the receiving end, ie, the physical layer.

[0134] Optionally, the receiving of the packet verification encryption semantic information sent by the sending end includes:

[0135] The received block encrypted semantic modulation information is subjected to signal demodulation and channel decoding to obtain the block verification encrypted semantic information.

[0136] By performing signal demodulation and channel decoding on the received packet encrypted semantic modulation information, the packet checksum encrypted semantic information can be restored and errors that may occur during the transmission process can be corrected, thus ensuring the accuracy and reliability of communication.

[0137] Specifically, first, the received information is demodulated to obtain a restored block encrypted semantic channel code; then, the restored block encrypted semantic channel code is correspondingly channel-decoded to obtain a restored block encrypted semantic frame.

[0138] Step 302: Based on information verification decoding processing, the block verification encryption semantic information is converted into block encryption semantic information.

[0139] This step can be applied to the second layer of the sending end, namely the network layer.

[0140] This step is used to verify information, identify bit errors in a timely manner, and flexibly choose whether to retransmit bit errors based on the accurate assessment of the importance of information at different levels and dimensions. This ensures the reliability of key information more intelligently during transmission and achieves more efficient and accurate information transmission.

[0141] The grouped verification encryption semantic information includes grouped encryption semantic information, a cyclic redundancy check code and an importance information check code.

[0142] Optionally, the information verification-based decoding process converts the group verification encryption semantic information into group encryption semantic information, including steps 401 to 403.

[0143] Step 401: Verify the importance information verification code.

[0144] The importance information check code is used to check the importance level of the received information and whether an error occurs in the transmission process of the importance indication information.

[0145] Step 402: If the importance information check code is correct, check the cyclic redundancy check code.

[0146] Optionally, in the case where the importance information check code is erroneous, an information retransmission request corresponding to the group verification encryption semantic information is sent.

[0147] In this step, the importance information check code is verified. Specifically, the recovered block encrypted semantic frame is checked against the check code at the end to determine whether the information is important. If the check code is incorrect, a retransmission request is immediately sent to determine the importance level of the information. After determining the importance level, further confirmation is performed to determine whether the information contains any errors.

[0148] Step 403: When the cyclic redundancy check code is correct, perform information decoding processing on the group check encryption semantic information to obtain the group encryption semantic information.

[0149] Under the premise of ensuring that the header importance indication information is correct, the cyclic redundancy check code is checked, that is, CRC check. The decision on whether to retransmit is made based on the header importance indication information and whether the CRC check is wrong.

[0150] Optionally, in the event that the cyclic redundancy check code is erroneous, the importance level of the group check encryption semantic information is determined to be higher than a preset first threshold based on the importance indication information; when the judgment result is yes, an information retransmission request corresponding to the group check encryption semantic information is sent.

[0151] Specifically, a first threshold is set based on actual application requirements. When the importance level is higher than the preset first threshold, the corresponding information is considered important. If a cyclic redundancy check error occurs, error retransmission is triggered to ensure the reliability of the important information. When the importance level is lower than the preset first threshold, the corresponding information is considered unimportant, and retransmission is determined based on the actual situation. This prioritizes the correct transmission of important information and optimizes system resource allocation.

[0152] Optionally, when the importance level is lower than a preset first threshold and the delay and transmission bandwidth of the communication system are within a preset second threshold range, an information retransmission request corresponding to the packet verification encryption semantic information is sent.

[0153] Specifically, the present disclosure prioritizes retransmission of information with high importance and errors. If the latency and bandwidth in this scenario are sufficient, more data is allowed to be retransmitted, and then the information with lower importance and errors is retransmitted in sequence. If latency and bandwidth do not allow for sufficient time, if the importance of the data is relatively low, retransmission is not performed. This ensures that the transmission of high-importance information is error-free, even if errors occur in less important information. For example, the important part of an image will remain clear, while the unimportant part with errors will appear blurred.

[0154] For information with correct importance information check code and cyclic redundancy check code, information decoding processing is performed to separate the packet header information and importance indication information from the block encryption semantic information, facilitating the next step of processing the block encryption semantic information.

[0155] Step 303: Decapsulate the block encrypted semantic information to obtain a block encrypted semantic vector.

[0156] This step can be applied to the third layer of the sending end, namely the transport layer.

[0157] The block encryption semantic information includes the block encryption semantic vector, importance indication information and packet header information. This step is used to separate the block encryption semantic vector from the packet header information and importance indication information so that the block encryption semantic vector can be further processed separately in subsequent steps.

[0158] Optionally, this step includes: performing information decoding processing on the grouped encrypted semantic information and discarding the importance indication information; and grouping the grouped encrypted semantic information according to the packet header information to obtain the grouped encrypted semantic vector.

[0159] Specifically, the header of the recovered block-encrypted semantic data packet is first removed, along with the importance indicator information. This information is then combined based on the position information in the packet header to obtain the recovered block-encrypted semantic vector. This process is also performed in groups. For example, all received data belonging to block-encrypted semantic vector 1 is combined to obtain recovered block-encrypted semantic vector 1, and the same is true for recovered block-encrypted semantic vector 2.

[0160] Step 304: perform semantic decryption and decoding processing on the block encrypted semantic vector to obtain source information.

[0161] This step can be applied to the fourth layer of the sending end, namely the application layer.

[0162] Optionally, step 304 includes steps 501 to 505 .

[0163] Step 501: Decrypt the group encrypted semantic vector to obtain a group semantic vector.

[0164] Step 502: reorganize semantic information of the grouped semantic vectors to obtain reorganized semantic information.

[0165] Step 503, the recombinant semantic information is spliced to obtain semantic coding information.

[0166] Step 504: Dequantize and decode the semantically coded information to obtain semantic features.

[0167] Step 505: Obtain the restored source information based on the semantic features.

[0168] For example, the recovered group encrypted semantic vectors are first decrypted to obtain recovered group semantic vectors. Semantic information is then reassembled from the recovered group semantic vectors to obtain recovered group semantic coded data. The recovered group semantic coded data are then concatenated to obtain recovered semantic coded data. The semantic coded data are then dequantized and source decoded to obtain recovered semantic features. Finally, semantic recovery is performed on the recovered semantic features to obtain recovered information.

[0169] Steps 501 to 505 correspond to steps 201 to 205, and their principles and functions are not repeated here.

[0170] The semantic communication method provided by the present disclosure ensures the security, reliability and efficiency of information transmission, optimizes the communication system structure and saves communication costs.

[0171] Fourth embodiment

[0172] Figure 7 This is one of the structural diagrams of the semantic communication device provided by the present disclosure. Figure 13 The semantic communication device 700 provided by the present disclosure is applied to a receiving end and may include:

[0173] The receiving module 701 is configured to receive the packet verification encryption semantic information sent by the sending end.

[0174] The check decoding module 702 is configured to convert the group check encryption semantic information into group encryption semantic information based on information check decoding processing.

[0175] The decapsulation module 703 is configured to decapsulate the block encrypted semantic information to obtain a block encrypted semantic vector.

[0176] The semantic decoding module 704 is configured to perform semantic decryption and decoding processing on the block encrypted semantic vector to obtain information source information.

[0177] The semantic communication device provided by the present disclosure ensures the security, reliability and efficiency of information transmission, optimizes the communication system structure and saves communication costs.

[0178] Fifth embodiment

[0179] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0180] Figure 14 A schematic block diagram of an example electronic device 800 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 laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are provided as examples only and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0181] like Figure 8 As shown, the device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. Various programs and data required for the operation of the device 800 can also be stored in the RAM 803. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0182] Various components in device 800 are connected to I / O interface 805, including an input unit 806, such as a keyboard, mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a magnetic disk, optical disk, etc.; and a communication unit 809, such as a network card, modem, wireless communication transceiver, etc. The communication unit 809 allows device 800 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0183] The computing unit 801 can be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 801 performs the various methods and processes described above, such as method XXX. For example, in some embodiments, method XXX can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as a storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on the device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of method XXX described above can be performed. Alternatively, in other embodiments, the computing unit 801 can be configured to perform method XXX in any other appropriate manner (e.g., by means of firmware).

[0184] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system comprising at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0185] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0186] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. A 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, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0187] To provide interaction with a user, the systems and techniques described herein 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 pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the 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 input, voice input, or tactile input).

[0188] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, 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.

[0189] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact through a communication network. The client-server relationship arises through computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, a server in a distributed system, or a server integrated with a blockchain.

[0190] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not limited herein.

[0191] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A semantic communication method, applied to a sending end, comprising: Performing encrypted semantic coding and weight configuration on the source information to obtain matching block encrypted semantic vectors and importance indication information; performing weighted coding processing on the matched block encryption semantic vector and the importance indication information to obtain block encryption semantic information; Based on information verification coding processing, the block encrypted semantic information is converted into block verification encrypted semantic information; The packet verification encryption semantic information is transmitted to a receiving end.

2. The method according to claim 1, wherein The encrypted semantic coding and weight configuration of the source information to obtain a matching grouped encrypted semantic vector and importance indication information includes: Extracting semantic features of the source information; Based on a preset weight allocation model, assigning weights to the semantic features according to the importance levels of the semantic features to obtain the importance indication information; performing source coding and quantization processing on the semantic features to obtain semantic coding information; performing grouping and vectorization processing on the semantic coding information to obtain a grouped semantic vector; encrypting the group semantic vector to obtain the group encrypted semantic vector; Each group of the group encryption semantic vectors matches the importance indication information.

3. The method according to claim 2, wherein: The preset weight distribution model includes but is not limited to: an information entropy importance distribution model and / or an importance experience distribution model.

4. The method according to claim 2, wherein: The grouping processing of the semantic coding information includes: setting the grouping granularity according to the application scenario.

5. The method according to claim 1, wherein The performing weighted coding processing on the matched block encryption semantic vector and the importance indication information to obtain block encryption semantic information includes: The matched block encryption semantic vector, the importance indication information, and the packet header information are encoded to obtain the block encryption semantic information.

6. The method according to claim 1, wherein The information verification encoding process is based on converting the block encrypted semantic information into block verification encrypted semantic information, including: Based on the importance indication information, the group encryption semantic information, the cyclic redundancy check code and the importance information check code are information-encoded to obtain the group check encryption semantic information.

7. The method according to claim 1, wherein The transmitting the packet verification encryption semantic information to the receiving end includes: Perform channel coding and signal modulation based on the block check encryption semantic information to obtain block encryption semantic modulation information; The transmission channel is matched according to the importance indication information, and the block encrypted semantic modulation information is transmitted to the receiving end.

8. A semantic communication device, applied to a sending end, comprising: The semantic coding module is used to perform encrypted semantic coding processing and weight configuration on the source information to obtain the matching block encrypted semantic vector and importance indication information; A weighted coding module, configured to perform weighted coding processing on the matched block encryption semantic vector and the importance indication information to obtain block encryption semantic information; A checksum coding module, configured to convert the block encrypted semantic information into block checksum encrypted semantic information based on information checksum coding processing; The sending module is used to transmit the group verification encryption semantic information to the receiving end.

9. A semantic communication method, applied to a receiving end, comprising: Receive the packet verification encryption semantic information sent by the sender; Based on information verification decoding processing, the group verification encryption semantic information is converted into group encryption semantic information; Decapsulating the block encrypted semantic information to obtain a block encrypted semantic vector; Perform semantic decryption and decoding processing on the block encrypted semantic vector to obtain source information.

10. The method according to claim 9, wherein: The receiving end sends encrypted semantic information of the packet verification, including: The received block encrypted semantic modulation information is subjected to signal demodulation and channel decoding to obtain the block verification encrypted semantic information.

11. The method according to claim 9, wherein The grouped verification encryption semantic information includes grouped encryption semantic information, a cyclic redundancy check code, and an importance information check code; The information verification decoding process is based on converting the group verification encryption semantic information into group encryption semantic information, including: Verifying the importance information verification code; If the importance information check code is correct, checking the cyclic redundancy check code; When the cyclic redundancy check code is correct, information decoding processing is performed on the group check encryption semantic information to obtain the group encryption semantic information.

12. The method according to claim 11, wherein The method further comprises: In the case that the importance information check code is wrong, an information retransmission request corresponding to the group verification encryption semantic information is sent.

13. The method according to claim 11, wherein The method further comprises: In the case where the cyclic redundancy check code is wrong, determining, based on the importance indication information, whether the importance level of the group verification encryption semantic information is higher than a preset first threshold; When the judgment result is yes, an information retransmission request corresponding to the packet verification encryption semantic information is sent.

14. The method according to claim 13, wherein The method further comprises: When the importance level is lower than a preset first threshold and the delay and transmission bandwidth of the communication system are within a preset second threshold range, an information retransmission request corresponding to the packet verification encryption semantic information is sent.

15. The method according to claim 9, wherein The block encryption semantic information includes a block encryption semantic vector, importance indication information and packet header information; The decapsulating the block encrypted semantic information to obtain the block encrypted semantic vector includes: performing information decoding processing on the block encrypted semantic information and discarding the importance indication information; The block encryption semantic information is grouped according to the packet header information to obtain the block encryption semantic vector.

16. The method according to claim 9, wherein The performing semantic decryption and decoding processing on the block encrypted semantic vector to obtain source information includes: Decrypting the grouped encrypted semantic vector to obtain a grouped semantic vector; Reorganizing semantic information of the grouped semantic vectors to obtain reorganized semantic information; splicing the recombined semantic information to obtain semantic coding information; performing dequantization processing and source decoding on the semantically coded information to obtain semantic features; Based on the semantic features, the restored source information is obtained.

17. A semantic communication device, applied to a receiving end, comprising: A receiving module, configured to receive the packet verification encrypted semantic information sent by the sending end; A checksum decoding module, configured to convert the group checksum encrypted semantic information into group encrypted semantic information based on information checksum decoding processing; a decapsulation module, configured to decapsulate the block encrypted semantic information to obtain a block encrypted semantic vector; The semantic decoding module is used to perform semantic decryption decoding processing on the block encrypted semantic vector to obtain source information.

18. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed 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 of any one of claims 1 to 7, or the method of any one of claims 9 to 16.

19. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to enable the computer to execute the method according to any one of claims 1 to 7, or to execute the method according to any one of claims 9 to 16.

20. A computer program product, comprising a computer program, wherein the computer program is configured to execute the method according to any one of claims 1 to 7 or the method according to any one of claims 9 to 16 when executed by a processor.

Images