Information transmission method and device, information processing method and device and semantic communication system

Abstract

The invention provides an information transmission method and device, and relates to the technical field of communication. According to the specific implementation scheme, semantic features of to-be-transmitted information in a current transmission scene are extracted; based on the current transmission scene and the semantic features, obtaining importance level information of a feature point group set comprising at least one feature point group; performing source-channel joint coding on the semantic features to obtain semantic coding data; based on the current transmission scene, grouping and vectorizing the semantic coded data to obtain a semantic vector group set comprising at least one semantic vector group, the semantic vector groups being in one-to-one correspondence with the feature point groups; obtaining a transmission semantic frame based on the importance level information and the semantic vector group set; and modulating the transmission semantic frame to obtain and adopt a wireless channel to transmit modulation information. According to the embodiment, the information transmission efficiency is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of communication technology, and more particularly to an information transmission method and apparatus, an information processing method and apparatus, a semantic communication system, an electronic device, and a computer-readable medium. Background Art

[0002] With the advent of the 6G era (6th Generation Mobile Networks), semantic communication, as one of the key technologies, plays a vital role. Unlike traditional communication technologies, semantic communication focuses on the "semantic similarity" of transmitted information rather than bit-level error-free transmission. This enables semantic communication to maintain good communication even in challenging communication environments.

[0003] However, semantic encoding and decoding as well as knowledge base updates consume a lot of computing and storage resources. How to effectively transmit semantic information is an urgent problem to be solved. Summary of the Invention

[0004] Provided are an information transmission method and device, an information processing method and device, a semantic communication system, an electronic device, and a computer-readable medium.

[0005] According to a first aspect, an information transmission method is provided, which includes: extracting semantic features of information to be transmitted in a current transmission scenario; obtaining importance level information of a feature point group set including at least one feature point group based on the current transmission scenario and the semantic features; performing source-channel joint encoding on the semantic features to obtain semantic coding data; based on the current transmission scenario, grouping and vectorizing the semantic coding data to obtain a semantic vector group set including at least one semantic vector group, wherein the semantic vector groups correspond one-to-one to the feature point groups; obtaining a transmission semantic frame based on the importance level information and the semantic vector group set; modulating the transmission semantic frame to obtain and transmit the modulation information using a wireless channel.

[0006] According to the second aspect, an information processing method is provided, which includes: demodulating the received modulated information to obtain a transmission semantic frame; verifying the transmission semantics; in response to the transmission semantic frame passing the verification, obtaining a set of semantic vector groups based on the transmission semantic frame; obtaining semantic coding data based on the set of semantic vector groups; performing source-channel joint decoding on the semantic coding data to obtain semantic features; and obtaining information to be transmitted based on the semantic features.

[0007] According to a third aspect, an information transmission device is provided, which includes: an extraction unit configured to extract semantic features of information to be transmitted in a current transmission scenario; a level acquisition unit configured to obtain important level information of a feature point group set including at least one feature point group based on the current transmission scenario and the semantic features; an encoding unit configured to perform source-channel joint encoding on the semantic features to obtain semantic coding data; a grouping unit configured to group and vectorize the semantic coding data based on the current transmission scenario to obtain a semantic vector group set including at least one semantic vector group, wherein the semantic vector group corresponds one-to-one to the feature point group; a frame acquisition unit configured to obtain a transmission semantic frame based on the important level information and the semantic vector group set; and a modulation unit configured to modulate the transmission semantic frame to obtain and transmit the modulation information using a wireless channel.

[0008] According to a fourth aspect, an information processing device is provided, which includes: a demodulation unit, configured to demodulate received modulated information to obtain a transmission semantic frame; a verification unit, configured to verify the transmission semantics; a semantics obtaining unit, configured to obtain a set of semantic vector groups based on the transmission semantic frame in response to passing the verification of the transmission semantic frame; a coding obtaining unit, configured to obtain semantic coding data based on the set of semantic vector groups; a decoding unit, configured to perform source-channel joint decoding on the semantic coding data to obtain semantic features; and an information obtaining unit, configured to obtain information to be transmitted based on the semantic features.

[0009] According to the fifth aspect, a semantic communication system is provided, which includes: a receiving end and a transmitting end for transmitting information through a wireless channel; the application layer of the transmitting end is used to extract the semantic features of the information to be transmitted in the current transmission scenario; based on the current transmission scenario and the semantic features, the importance level information of the feature point group set including at least one feature point group is obtained; the semantic features are jointly coded with the source channel to obtain semantic coded data; wherein the channel state information used in the source channel joint coding is obtained from the physical layer of the transmitting end; the transmission and network layers of the transmitting end are used to group and vectorize the semantic coded data based on the current transmission scenario to obtain a semantic vector group set including at least one semantic vector group, wherein the semantic vector group corresponds one-to-one to the feature point group; each semantic vector group in the semantic vector group set is segmented to obtain a plurality of segmented sub-vectors; based on the importance level information, a packet is added to each segmented sub-vector in the plurality of segmented sub-vectors. header information, and obtain multiple data packets with importance levels; the data link layer at the sending end generates a check code for the data packet corresponding to each segmentation subvector in the semantic vector group set based on the importance level information of the data packet, and adds the check code of the data packet to the end of the data packet to obtain a transmission semantic frame of the data packet; the physical layer at the sending end is used to modulate the transmission semantic frame to obtain and transmit the modulation information through the wireless channel; the physical layer at the receiving end is used to demodulate the modulation information received from the wireless channel to obtain a transmission semantic frame; the data link layer at the receiving end is used to verify the transmission semantics; the transmission and network layer at the receiving end is used to obtain a semantic vector group set based on the transmission semantic frame in response to the transmission semantic frame verification; based on the semantic vector group set, obtain semantic coding data; the application layer at the receiving end performs source-channel joint decoding on the semantic coding data to obtain semantic features; based on the semantic features, obtain information to be transmitted.

[0010] According to the sixth aspect, an electronic device is provided, which includes: 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 method described in any implementation of the first aspect or the second aspect.

[0011] According to a seventh aspect, a non-transitory computer-readable storage medium storing computer instructions is provided, where the computer instructions are used to cause a computer to execute the method described in any implementation of the first aspect or the second aspect.

[0012] The information transmission method and device provided by the embodiments of the present disclosure first extract the semantic features of the information to be transmitted in the current transmission scenario; secondly, based on the current transmission scenario and the semantic features, obtain the importance level information of the feature point group set including at least one feature point group; thirdly, perform source-channel joint coding on the semantic features to obtain semantic coding data; thirdly, based on the current transmission scenario, group and vectorize the semantic coding data to obtain a semantic vector group set including at least one semantic vector group, wherein the semantic vector group corresponds one-to-one with the feature point group; then, based on the importance level information and the semantic vector group set, obtain the transmission semantic frame; finally, modulate the transmission semantic frame to obtain and adopt the wireless channel to transmit the modulation information. Thus, by extracting the semantic features, the information is compressed at the semantic level, and the amount of information to be transmitted is reduced; by using the source-channel joint coding, the channel information can be effectively combined while retaining the relevance of the semantic information; by characterizing the importance of the semantic information through the importance level information, it is ensured that the transmission of information with high importance level does not cause errors, thereby improving the reliability of information transmission.

[0013] 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

[0014] The accompanying drawings are used to better understand the present invention and do not constitute a limitation of the present invention.

[0015] Figure 1 is a flow chart of an embodiment of the information transmission method according to the present disclosure;

[0016] Figure 2 is a flow chart of an embodiment of an information processing method according to the present disclosure;

[0017] Figure 3 is a schematic structural diagram of an embodiment of an information transmission device according to the present disclosure;

[0018] Figure 4 is a schematic structural diagram of an embodiment of an information processing device according to the present disclosure;

[0019] Figure 5 is a structural diagram of an embodiment of a semantic communication system according to the present disclosure;

[0020] Figure 6 It is a block diagram of an electronic device used to implement the information transmission method or information processing method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0021] 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.

[0022] In this embodiment, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features.

[0023] This disclosure proposes an information transmission method. Figure 1 A process 100 according to an embodiment of the information transmission method of the present disclosure is shown. The information transmission method includes the following steps:

[0024] Step 101: extracting semantic features of information to be transmitted in the current transmission scenario.

[0025] In this embodiment, the information to be transmitted is the information required to be transmitted according to the present disclosure. The information to be transmitted can be text data, image data, or voice data. The current transmission scenario can be the scenario corresponding to the information to be transmitted, such as gaming scenarios, autonomous driving, satellite communications, the Internet of Things, and other scenarios. The content of the information to be transmitted varies for different transmission scenarios. For example, in gaming scenarios, the information to be transmitted includes information such as the player's actions, the game environment, and the game status. The information transmission method disclosed herein can effectively reduce information transmission delay. In autonomous driving scenarios, the information to be transmitted can include information such as the vehicle's location, speed, and direction of travel. The information transmission method disclosed herein can transmit critical information for autonomous driving even under poor channel quality, ensuring that important information is not lost. In satellite communications scenarios, the information to be transmitted can include emergency signals transmitted via voice, such as distress signals, location information, injury status, and other information. The information transmission method disclosed herein can ensure that even under poor signal quality, the most important information is received. For the Internet of Things scenario, the information to be transmitted can be: transmission data from sensors and devices in text form. Through the information transmission method disclosed in the present invention, text data can be intelligently processed and encoded, reducing the transmission of irrelevant and redundant information, improving the overall efficiency of the network, and ensuring the reliability of the transmission of key information in various environments.

[0026] Step 102: Obtain importance level information of a feature point group set including at least one feature point group based on the current transmission scenario and semantic features.

[0027] In this embodiment, the importance level information is information representing the importance of different features in the semantic features. For example, the importance level information includes: level values, such as level 1, level 2...

[0028] The above-mentioned step 102 includes: based on the current transmission scenario, grouping the feature points in the semantic feature to obtain a feature point group set including at least one feature point group; based on the influence of each feature point group on the current transmission scenario, dividing each feature resistor into different levels to obtain important level information of the feature point group set.

[0029] The information to be transmitted will be converted into digital data for transmission. After semantic extraction of the original information to be transmitted, semantic features are obtained. The feature point refers to each point in the semantic feature. The information content contained in this feature point is a continuous numerical information, and the occurrence of this numerical value has a certain probability (similar to the symbol probability in traditional communication technology).

[0030] Step 103: Perform source-channel joint coding on the semantic features to obtain semantic coding data.

[0031] In this embodiment, semantic features are jointly encoded with channel state information to obtain semantically encoded data. The source-channel joint encoding requires the underlying physical layer to upload the channel state information to the application layer so that the encoded semantically encoded data can adapt to the channel. The transmission process can be described as follows: the physical layer or data link layer detects the channel state information, encapsulates the channel state information into data packets at the network layer and transport layer, and transmits it to the application layer. The application layer extracts the detected channel state information and utilizes it in the source-channel joint encoding. The uploading of channel state information also requires a certain mechanism, and the channel state information can change dramatically over time.

[0032] Step 104 : Based on the current transmission scenario, the semantically coded data is grouped and vectorized to obtain a semantic vector group set including at least one semantic vector group.

[0033] In this embodiment, the semantic vector groups correspond to the feature point groups one by one.

[0034] In this embodiment, when grouping semantically coded data, reference can be made to the impact of the semantically coded data on the current transmission scenario. For example, different granularities can be set for grouping. For the current transmission scenario, data in the semantically coded data that is suitable for semantic communication enhancement scenarios can be divided into higher granularities, and data in the semantically coded data that is suitable for semantic communication empowerment scenarios can be divided into lower granularities.

[0035] Step 105: Obtain a transmission semantic frame based on the importance level information and the semantic vector group set.

[0036] In this embodiment, the transmission semantic frame is a protocol data unit of the data link layer. The transmission semantic frame includes three parts: a frame header, a data part, and a frame trailer. The frame header and frame trailer contain some necessary control information, such as synchronization information, address information, and error control information; the data part contains data transmitted from the network layer.

[0037] In this embodiment, a frame header corresponding to the semantic vector group set is generated based on the importance level information, and the frame header is appended to the front of the data packet pre-obtained from the semantic vector group set to obtain a transmission semantic frame. Alternatively, a frame trailer corresponding to the semantic vector group set is generated based on the importance level information, and the frame trailer is appended to the back of the data packet pre-obtained from the semantic vector group set to obtain a transmission semantic frame.

[0038] Step 106: modulate the transmission semantic frame to obtain and transmit modulation information via a wireless channel.

[0039] In this embodiment, the transmission semantic frame is modulated to obtain modulation information, and the modulation information is sent to the wireless channel for transmission, wherein the modulation method of the modulation information will adaptively select the optimal modulation scheme according to the state of the channel.

[0040] When selecting different modulation methods, the physical layer determines the most appropriate modulation scheme based on channel state information (such as channel quality, signal-to-noise ratio, and channel bandwidth). This process is often called adaptive modulation. For example, when the link quality between the transmitter and receiver is excellent, the system may use higher-order modulation (such as 64-QAM or higher) to achieve higher data throughput. In weaker signal conditions, QPSK (Quadrature Phase Shift Keying) or lower-order modulation methods may be used to ensure reliable signal transmission.

[0041] The information transmission method provided by the embodiment of the present disclosure first extracts the semantic features of the information to be transmitted in the current transmission scenario; secondly, based on the current transmission scenario and the semantic features, obtains the importance level information of the feature point group set including at least one feature point group; thirdly, performs source-channel joint coding on the semantic features to obtain semantic coding data; thirdly, based on the current transmission scenario, the semantic coding data is grouped and vectorized to obtain a semantic vector group set including at least one semantic vector group, wherein the semantic vector group corresponds one-to-one to the feature point group; then, based on the importance level information and the semantic vector group set, a transmission semantic frame is obtained; finally, the transmission semantic frame is modulated to obtain and adopt a wireless channel to transmit the modulation information. Thus, by extracting semantic features, information compression at the semantic level is achieved, reducing the amount of information to be transmitted; by using source-channel joint coding, the channel information can be effectively combined while retaining the relevance of semantic information; the importance of semantic information is represented by the importance level information, thereby improving the reliability of information transmission.

[0042] In some embodiments of the present disclosure, the information transmission method further includes: encrypting each semantic vector group in the semantic vector group set using an encryption algorithm.

[0043] In this embodiment, for the current transmission scenario, the semantic vector group set may include: a high-grained semantic vector group and a low-grained semantic vector group, wherein the high-grained semantic vector group is obtained through the high-grained semantic coding data group under the current transmission scenario (for example, an image transmission scenario or a text transmission scenario), and the low-grained semantic vector group is obtained through the low-grained semantic coding data group under the current transmission scenario.

[0044] The above-mentioned encryption of each semantic vector group in the semantic vector group set using an encryption algorithm includes: encrypting the high-grained semantic vector group and the low-grained semantic vector group using an encryption algorithm respectively, and the data of the high-grained semantic vector group and the low-grained semantic vector group do not interfere with each other, that is, the encryption process of the high-grained semantic vector group will not be interfered with by other semantic vector groups, and the encryption process of the low-grained semantic vector group will not be interfered with by other semantic vector groups.

[0045] The information transmission method provided in this embodiment encrypts each semantic vector in the semantic vector group set after performing source-channel joint encoding on the semantic features, so that the encryption does not affect the correlation of the information after the source-channel joint encoding, thereby ensuring the integrity of the information.

[0046] In some embodiments of the present disclosure, the above-mentioned obtaining of importance level information of a feature point group set including at least one feature point group based on the current transmission scenario and semantic features includes: calculating the information entropy of each feature point in the semantic feature; grouping the information entropy of the feature points based on the current transmission scenario to obtain a feature point group set of at least one feature point group; calculating the normalized average value of the information entropy of each feature point group in the feature point group set; and obtaining the importance level information of each feature point group in the feature point group set based on the normalized average value.

[0047] In this embodiment, the above-mentioned calculation of the normalized average value of the information entropy of each feature point group in the feature point group set includes: obtaining the normalized value of the information entropy of each feature point in each feature point group based on the maximum value, minimum value of the information entropy of each feature point group in the feature point group set and the information entropy of each feature point; averaging the normalized values of all the information entropies in each feature point group to obtain the normalized average value of the information entropy of each feature point group.

[0048] In this embodiment, the importance level of a feature point with a large information entropy is higher, and the importance level of a feature point with a small information entropy is lower. Since the maximum value of the information entropy of different information is uncertain, this disclosure adopts the information entropy normalization value, assuming that the maximum value of the information entropy of the current semantic feature is max and the minimum value is min, and assuming that the information entropy value of each feature point is e ij ,pass The information entropy is normalized to between 0 and 1 to obtain the normalized value of the information entropy.

[0049] At the same time, the normalized values of the obtained information entropy are grouped in the same manner according to the grouping method of the semantically coded data in the current transmission scenario, and the normalized values of the grouped information entropy are averaged to obtain the normalized average value.

[0050] The method for obtaining importance level information provided in this embodiment calculates the information entropy of each feature point in the semantic feature, groups the information entropy to obtain a feature point group set, calculates the normalized average value of the information entropy of each feature point group in the feature point group set, and obtains the importance level information of each feature point group based on the normalized average value, thereby providing a reliable implementation method for obtaining the importance level information.

[0051] In some optional implementations of this embodiment, the above-mentioned obtaining of the importance level information of each feature point group in the feature point group set based on the normalized average value includes: based on the normalized average value of the information entropy of each feature point group in the feature point group set, querying a pre-set importance mapping set to obtain the importance level information of each feature point group, and the importance mapping set is used to characterize the correspondence between the normalized average value and the importance level information.

[0052] In this embodiment, after obtaining the normalized value of the information entropy of each feature point group, a level of information importance is assigned to each feature point group according to a preset importance mapping set to identify the importance level information, for example:

[0053] When the data value of the normalized value of the information entropy of the feature point group belongs to [0, 0.2), the importance level information is set to level 1 (specific different levels may be represented by different data); when the data value of the normalized value of the information entropy of the feature point group belongs to (0.2, 0.4], the importance level information is set to level 2; and so on, when the data value of the normalized value of the information entropy of the feature point group belongs to (0.8, 1], the importance level information is set to level 5, and level 5 indicates that the current information is the most important information.

[0054] The method for obtaining the importance level information of the feature point group provided in this embodiment obtains the importance level information of the feature point group through the normalized average value of the information entropy of each feature point group in the feature point group set and a pre-set importance mapping set, providing a reliable implementation method for obtaining the importance level information.

[0055] Optionally, obtaining the importance level information of each feature point group in the feature point group set based on the normalized average value includes: inputting the normalized average value of the information entropy of each feature point group in the feature point group set into a pre-trained importance level assessment model, and obtaining the importance level information of each feature point group in the feature point group set output by the importance level assessment model. The importance level assessment model is used to characterize the correspondence between the normalized average value of the information entropy and the importance level information.

[0056] In some optional implementations of this embodiment, the above-mentioned importance level information of the feature point group set including at least one feature point group based on the current transmission scenario and semantic features includes: grouping the feature points in the semantic features based on the current transmission scenario to obtain a feature point group set including at least one feature point group; and obtaining the importance level information of each feature point group in the feature point group set based on the information entropy of the feature points in the semantic features.

[0057] The method for obtaining the importance level information of a feature point group set provided in this embodiment first groups the feature points in the semantic feature to obtain a feature point group set; based on the information entropy of the feature points in the semantic feature, the importance level information of each feature point group in the feature point group set is obtained, providing another reliable implementation method for obtaining the importance level information.

[0058] In some optional implementations of this embodiment, the above-mentioned grouping and vectorization processing of the semantically coded data based on the current transmission scenario to obtain a semantic vector group set including at least one semantic vector group includes: based on the current transmission scenario, dividing the semantically coded data into semantically coded data groups belonging to high granularity and low granularity respectively; vectorizing the semantically coded data groups to obtain a semantic vector group set including at least one semantic vector group.

[0059] In this optional implementation, each semantic vector in the semantic vector group set may be an encrypted vector.

[0060] In this optional implementation, the current transmission scenario may include: an image data transmission scenario and a text data transmission scenario.

[0061] For image data transmission scenarios, the above-mentioned semantic coding data can be three-dimensional image semantic coding data. The above-mentioned division of the semantic coding data into semantic coding data groups belonging to high-granularity and low-granularity based on the current transmission scenario includes: for C×H×W three-dimensional image semantic coding data (where C is the number of data channels, H is the image height, and W is the image width), it can be divided into 2×2=4 areas of low-granularity semantic coding data groups, or 4×4=16 areas of high-granularity semantic coding data groups in the H×W dimension.

[0062] For text data transmission scenarios, the above-mentioned semantic coding data can be text semantic coding data. The above-mentioned division of semantic coding data into semantic coding data groups belonging to high-granularity and low-granularity based on the current transmission scenario includes: 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 semantic coding data groups with 100 coding data in each group, or high-granularity semantic coding data groups with 10 coding data in each group.

[0063] The method for obtaining a semantic vector group set provided in this embodiment can ensure that, during the transmission process of the semantic frame corresponding to the semantic vector group set, information with a high importance level will not be transmitted error-free. Even if an error occurs in information with a low importance level, for example, the important part of the image will still be clear, while the unimportant part with the error will appear blurred.

[0064] The method for obtaining a semantic vector group set provided in this embodiment divides the semantic coding data into high-granularity and low-granularity semantic coding data groups based on the current transmission scenario, which can effectively group the semantic vector groups in the semantic vector group set and provide a reliable implementation means for obtaining the important level information of each semantic vector group.

[0065] In some optional implementations of the present disclosure, the above-mentioned obtaining of a transmission semantic frame based on importance level information and a semantic vector group set includes: dividing each semantic vector group in the semantic vector group set to obtain a plurality of segmented subvectors; adding header information to each segmented subvector in the plurality of segmented subvectors based on the importance level information to obtain a plurality of data packets with importance level information; for a data packet corresponding to each segmented subvector in the semantic vector group set, generating a check code for the data packet based on the importance level information of the data packet, and adding the check code of the data packet to the tail of the data packet to obtain a transmission semantic frame of the data packet.

[0066] In this optional implementation, the semantic information of a semantic vector group cannot be effectively determined due to the length of the data segment. Therefore, the semantic vector group is segmented to effectively reflect the semantic meaning of the semantic vector. It should be noted that the segmented subvectors belonging to the same semantic vector group have the same importance level information. Therefore, when generating data packets, data packets belonging to the same semantic vector group have the same importance level information.

[0067] In this optional implementation, each semantic vector in the semantic vector group set may be an encrypted vector.

[0068] In this optional implementation, the packet header information includes: source address, destination address and importance level information.

[0069] In this optional implementation, the importance level information in each data packet is related to the semantic vector group. For example, after semantic vector group 1 is split to obtain multiple split sub-vectors, the source address and destination address are added. When adding the importance level information, all split sub-vectors from semantic vector group 1 will have the same importance level information added, and so on for semantic vector group 2.

[0070] In this optional implementation, the check code may be a cyclic redundancy check code. For the data packet corresponding to each segmented subvector in the semantic vector group set, the check code for the data packet is generated based on the importance level information of the data packet, and the check code for the data packet is added to the end of the data packet to obtain the transmission semantic frame of the data packet. The steps include: generating a cyclic redundancy check code (CRC code) for the data packet based on the importance level information of the data packet, and adding the CRC code to the end of the data packet to obtain the transmission semantic frame of the data packet. The cyclic redundancy check code is the most commonly used error checking code in the field of data communications, characterized in that the lengths of the information field and the check field can be arbitrarily selected. In this embodiment, the lengths of the information field and / or the check field of the cyclic redundancy check code can be generated based on the importance level information. For example, if the importance level of a data packet is high, the number of bits of its CRC check code will be longer or the number of CRC checks will be greater; if the importance level information of a data packet is low, the number of bits of its CRC check code will be shorter or the number of CRC checks will be less, thereby ensuring that the probability of errors occurring in more important data is lower.

[0071] This optional implementation provides a method for obtaining a transmission semantic frame by segmenting each semantic vector group in a set of semantic vector groups to obtain multiple segmented subvectors. Based on importance level information, header information is added to each of the multiple segmented subvectors to obtain multiple data packets with importance level information. Based on the importance level information of the data packets, a checksum is generated for the data packets and added to the end of the data packets to obtain the transmission semantic frame of the data packets. This provides a reliable implementation method for obtaining the transmission semantic frame of the semantic vector group.

[0072] Optionally, the above-mentioned method of obtaining a transmission semantic frame based on the importance level information and the semantic vector group set includes: dividing the semantic vectors in each semantic vector group of the semantic vector group set, and adding header information to the semantic vectors obtained by dividing the semantic vector group set based on the importance level information, to obtain multiple data packets with importance level information; for the data packet corresponding to each semantic vector, generating a check code for the data packet based on the importance level information of the data packet, and adding the check code of the data packet to the end of the data packet to obtain a transmission semantic frame of the data packet.

[0073] In some optional implementations of the present disclosure, for a data packet corresponding to each slice subvector in the semantic vector group set, generating a check code for the data packet based on importance level information of the data packet includes any of the following:

[0074] For the data packets corresponding to each segmentation subvector in the semantic vector group set, based on the importance level information of the data packet, a first cyclic redundancy check code of the number of bits corresponding to the importance level information of the data packet is generated; based on the importance level information of the data packet, a first level check code is generated, and the first cyclic redundancy check code and the first level check code are used as the check code of the data packet.

[0075] For the data packets corresponding to each segmentation subvector in the semantic vector group set, based on the importance level information of the data packet, a second cyclic redundancy check code with a check number corresponding to the importance level information of the data packet is generated; based on the importance level information of the data packet, a second level check code is generated, and the second cyclic redundancy check code and the second level check code are used as the check code of the data packet.

[0076] In this optional implementation, the first cyclic redundancy check code and the second cyclic redundancy check code are both CRC codes generated based on the importance level information. The first-level check code and the second-level check code are check codes used to verify the importance level information. The first-level check code and the second-level check code can be codes obtained by performing operations (such as convolution operations and cyclic operations) on the importance level information.

[0077] In this optional implementation, the above-mentioned generation of the first cyclic redundancy check code of the number of bits corresponding to the importance level information of the data packet based on the importance level information of the data packet includes: based on the importance level information of the data packet, querying a pre-set first-level cyclic redundancy check code correspondence table to obtain the first cyclic redundancy check code of the data packet. The first-level cyclic redundancy check code correspondence table is a table of correspondences between importance level information and cyclic redundancy check codes of different bits. In the first-level cyclic redundancy check code correspondence table, different importance level information corresponds to different cyclic redundancy check code bits. For example, CRC32 (i.e., a 32-bit cyclic redundancy check code) is used for importance level information of level 5; CRC24 is used for importance level information of level 4; CRC16 is used for importance level information of level 3; CRC12 is used for importance level information of level 2; and CRC8 is used for information of level 1.

[0078] In this optional implementation, the generating of the second cyclic redundancy check code having a number of bits corresponding to the importance level information of the data packet based on the importance level information of the data packet includes: querying a pre-set second-level cyclic redundancy check code correspondence table based on the importance level information of the data packet to obtain the second cyclic redundancy check code of the data packet. The second-level cyclic redundancy check code correspondence table is a table of correspondences between importance level information and different numbers of cyclic redundancy check codes. In the second-level cyclic redundancy check code correspondence table, different numbers of cyclic redundancy checks are performed on information of different importance levels. For example, four CRC checks are performed for information of level 4; three CRC checks are performed for information of level 3; two CRC checks are performed for information of level 2; and one CRC check is performed for information of level 1.

[0079] In this optional implementation, the generating of the first level check code based on the importance level information of the data packet includes: performing convolution processing on the importance level information of the data packet to obtain the first level check code.

[0080] In this optional implementation, the generating of the second level check code based on the importance level information of the data packet includes: performing convolution processing on the importance level information of the data packet to obtain the second level check code.

[0081] The method for generating a check code for a data packet provided by this optional implementation generates a first cyclic redundancy check code or a second cyclic redundancy check code based on the importance level information of the data packet; generates a first-level check code or a second-level check code based on the importance level information of the data packet, uses the first cyclic redundancy check code and the first-level check code as the check code of the data packet, or uses the second cyclic redundancy check code and the second-level check code as the check code of the data packet, providing a reliable implementation method for obtaining the check code of the data packet.

[0082] Optionally, for each data packet corresponding to each segmented subvector in the semantic vector group set, generating a check code for the data packet based on the importance level information of the data packet may further include: employing different check combinations according to the importance level information. For example, employing more check methods (parity check, checksum) for higher-level information.

[0083] This disclosure proposes an information processing method. Figure 2 A process 200 according to an embodiment of the information processing method of the present disclosure is shown. The information processing method includes the following steps:

[0084] Step 201: demodulate the received modulation information to obtain a transmission semantic frame.

[0085] In this embodiment, the received information is demodulated and restored to obtain a transmission semantic frame.

[0086] Step 202: Verify the transmission semantic frame.

[0087] In this embodiment, the above step 202 includes: checking the importance level information of the transmission semantic frame, and immediately retransmitting if the importance level information is determined to be incorrect through the check code of the frame header or frame tail of the transmission semantic frame; if the importance level information is correct, proceed to step 203.

[0088] Verifying the importance level of transmission semantic frames includes performing a CRC check on the transmission semantic frames. If the importance level information in the frame header is verified to be correct, retransmission is determined based on the importance level information in the header and whether the CRC check results in errors. First, retransmission of high-importance information with errors is guaranteed. If there is sufficient latency and bandwidth in this scenario, more data is allowed to be retransmitted, followed by retransmission of information with lower importance levels with errors.

[0089] Step 203: In response to the transmission semantic frame passing the check, a semantic vector group set is obtained based on the transmission semantic frame.

[0090] The frame header of the data packet corresponding to the transmission semantic frame is removed, and the importance level information is removed.

[0091] The above information is combined according to the position information in the packet header to obtain a restored semantic vector group set, and the semantic vectors in the semantic vector group are divided at the same time.

[0092] Step 204: Obtain semantic coding data based on the semantic vector group set.

[0093] In this embodiment, the restored semantic vectors are subjected to semantic information reorganization to obtain restored semantic coding data groups, and the restored semantic coding data groups are reassembled in groups to obtain restored semantic coding data.

[0094] Step 205: perform source-channel joint decoding on the semantically coded data to obtain semantic features.

[0095] In this embodiment, source-channel joint decoding is performed on the restored semantically coded data in combination with the channel state information uploaded from the physical layer to the application layer to obtain restored semantic features.

[0096] Step 206: Obtain information to be transmitted based on the semantic features.

[0097] In this embodiment, semantic restoration is performed on the restored semantic features to obtain restored information to be transmitted.

[0098] The information processing method provided by the embodiments of the present disclosure first demodulates the received modulated information to obtain a transmission semantic frame; secondly, verifies the transmission semantic frame; thirdly, in response to the verification of the transmission semantic frame, obtains a set of semantic vector groups; thirdly, obtains semantically coded data based on the set of semantic vector groups; thirdly, performs source-channel joint decoding on the semantically coded data to obtain semantic features; and finally, obtains the information to be transmitted based on the semantic features. Thus, the importance of semantic information is represented by importance level information, improving the reliability of information processing.

[0099] The semantic vector groups in the semantic vector group set may be encrypted data. In some embodiments of the present disclosure, the method further includes: decrypting each semantic vector group in the semantic vector group set.

[0100] Therefore, after obtaining the semantic vector group set, the semantic vector groups in the semantic vector group set can be decrypted, thereby improving the security of information processing.

[0101] Further references Figure 3 As an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of an information transmission device, which is similar to Figure 1 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0102] like Figure 3 As shown, the information transmission device 300 provided by this embodiment includes: an extraction unit 301, a level obtaining unit 302, an encoding unit 303, a grouping unit 304, a frame obtaining unit 305, and a modulation unit 306.

[0103] Among them, the above-mentioned extraction unit 301 can be configured to extract the semantic features of the information to be transmitted in the current transmission scenario. The above-mentioned level acquisition unit 302 can be configured to obtain the importance level information of the feature point group set including at least one feature point group based on the current transmission scenario and the semantic features. The above-mentioned encoding unit 303 can be configured to perform source-channel joint encoding on the semantic features to obtain semantic coding data. The above-mentioned grouping unit 304 can be configured to group and vectorize the semantic coding data based on the current transmission scenario to obtain a semantic vector group set including at least one semantic vector group, wherein the semantic vector group corresponds one-to-one to the feature point group. The above-mentioned frame acquisition unit 305 can be configured to obtain a transmission semantic frame based on the importance level information and the semantic vector group set. The above-mentioned modulation unit 306 can be configured to modulate the transmission semantic frame to obtain and transmit the modulation information using a wireless channel.

[0104] In this embodiment, the specific processing and technical effects of the extraction unit 301, the level obtaining unit 302, the encoding unit 303, the grouping unit 304, the frame obtaining unit 305, and the modulation unit 306 in the information transmission device 300 can be referred to respectively. Figure 1 The relevant descriptions of step 101, step 102, step 103, step 104, step 105, and step 106 in the corresponding embodiment are not repeated here.

[0105] In some optional implementations of this embodiment, the apparatus 300 further includes an encryption unit (not shown in the figure). The encryption unit may be configured to encrypt each semantic vector group in the set of semantic vector groups using an encryption algorithm.

[0106] In some optional implementations of this embodiment, the above-mentioned level obtaining unit 302 is further configured to: calculate the information entropy of each feature point in the semantic feature; group the information entropy of the feature points based on the current transmission scenario to obtain a feature point group set of at least one feature point group; calculate the normalized average value of the information entropy of each feature point group in the feature point group set; and obtain the importance level information of each feature point group in the feature point group set based on the normalized average value.

[0107] In some optional implementations of this embodiment, the above-mentioned level obtaining unit 302 is further configured to: based on the normalized average value of the information entropy of each feature point group in the feature point group set, query a pre-set importance mapping set to obtain the importance level information of each feature point group, and the importance mapping set is used to characterize the correspondence between the normalized average value and the importance level information.

[0108] In some optional implementations of this embodiment, the above-mentioned level obtaining unit 302 is configured to: group the feature points in the semantic feature based on the current transmission scenario to obtain a feature point group set including at least one feature point group; and obtain importance level information of each feature point group in the feature point group set based on the information entropy of the feature points in the semantic feature.

[0109] In some optional implementations of this embodiment, the above-mentioned grouping unit 304 is configured to: divide the semantically coded data into semantically coded data groups belonging to high granularity and low granularity respectively based on the current transmission scenario; and perform vectorization processing on the semantically coded data groups to obtain a semantic vector group set including at least one semantic vector group.

[0110] In some optional implementations of this embodiment, the frame obtaining unit 305 is configured to: divide each semantic vector group in the semantic vector group set to obtain multiple segmented subvectors; based on the importance level information, add header information to each segmented subvector in the multiple segmented subvectors to obtain multiple data packets with importance level information; for the data packet corresponding to each segmented subvector in the semantic vector group set, based on the importance level information of the data packet, generate a check code for the data packet, and add the check code of the data packet to the end of the data packet to obtain a transmission semantic frame of the data packet.

[0111] In some optional implementations of this embodiment, the frame obtaining unit 305 is further configured to implement any one of the following: for a data packet corresponding to each slice subvector in the semantic vector group set, generating, based on the importance level information of the data packet, a first cyclic redundancy check code having a number of bits corresponding to the importance level information of the data packet; generating a first-level check code based on the importance level information of the data packet, and using the first cyclic redundancy check code and the first-level check code as a check code for the data packet;

[0112] For the data packets corresponding to each segmentation subvector in the semantic vector group set, based on the importance level information of the data packet, a second cyclic redundancy check code with a check number corresponding to the importance level information of the data packet is generated; based on the importance level information of the data packet, a second level check code is generated, and the second cyclic redundancy check code and the second level check code are used as the check code of the data packet.

[0113] The information transmission device provided by the embodiment of the present disclosure is as follows: first, the extraction unit 301 extracts the semantic features of the information to be transmitted in the current transmission scenario; secondly, the level obtaining unit 302 obtains the importance level information of the feature point group set including at least one feature point group based on the current transmission scenario and the semantic features; thirdly, the encoding unit 303 performs source-channel joint coding on the semantic features to obtain semantic coding data; thirdly, the grouping unit 304 groups and vectorizes the semantic coding data based on the current transmission scenario to obtain a semantic vector group set including at least one semantic vector group, wherein the semantic vector group corresponds one-to-one to the feature point group; then, the frame obtaining unit 305 obtains a transmission semantic frame based on the importance level information and the semantic vector group set; finally, the modulation unit 306 modulates the transmission semantic frame to obtain and transmit the modulation information using a wireless channel. Thus, by extracting the semantic features, information compression at the semantic level is achieved, reducing the amount of information required to be transmitted; by using source-channel joint coding, channel information can be effectively combined while retaining the relevance of the semantic information; the importance of the semantic information is represented by the importance level information, thereby improving the reliability of information transmission.

[0114] Continue to see Figure 4, as a response to the above Figure 2 The present disclosure provides an embodiment of an information processing device. Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0115] like Figure 4 As shown, the information processing device 400 of this embodiment may include: a demodulation unit 401, a verification unit 402, a semantics obtaining unit 403, an encoding obtaining unit 404, a decoding unit 405, and an information obtaining unit 406.

[0116] Among them, the above-mentioned demodulation unit 401 can be configured to demodulate the received modulated information to obtain a transmission semantic frame. The above-mentioned verification unit 402 can be configured to verify the transmission semantics. The above-mentioned semantic acquisition unit 403 is configured to obtain a set of semantic vector groups based on the transmission semantic frame in response to the transmission semantic frame verification being passed. The above-mentioned coding acquisition unit 404 can be configured to obtain semantic coding data based on the set of semantic vector groups. The above-mentioned decoding unit 405 can be configured to perform source-channel joint decoding on the semantic coding data to obtain semantic features. The above-mentioned information acquisition unit 406 is configured to obtain information to be transmitted based on the semantic features.

[0117] In this embodiment, the specific processing of the demodulation unit 401, the verification unit 402, the semantic obtaining unit 403, the encoding obtaining unit 404, the decoding unit 405, and the information obtaining unit 406 and the technical effects thereof can be referred to respectively. Figure 2 The relevant descriptions of step 201, step 202, step 203, step 204, step 205, and step 206 in the corresponding embodiment are not repeated here.

[0118] In some optional implementations of this embodiment, the apparatus 400 further includes a decryption unit (not shown in the figure). The decryption unit may be configured to decrypt each semantic vector group in the set of semantic vector groups.

[0119] Continue to see Figure 5 The present disclosure also provides a semantic communication system 500 , which includes: a sending end 502 and a receiving end 503 for transmitting information through a wireless channel 501 .

[0120] The application layer of the transmitter 502 is configured to extract semantic features of the information to be transmitted in the current transmission scenario; based on the current transmission scenario and the semantic features, obtain importance level information for a feature point group set including at least one feature point group; and perform source-channel joint coding on the semantic features to obtain semantically coded data. The channel state information used in the source-channel joint coding is obtained from the physical layer of the transmitter. For a detailed description of the functions of the application layer of the transmitter 502, see steps 101 to 103 of the illustrated embodiment of the information transmission method.

[0121] Joint source-channel coding requires the underlying physical layer to upload channel state information to the application layer, allowing the encoded semantically coded data to adapt to the channel. The transmission process can be described as follows: the physical layer or data link layer detects the channel state information, encapsulates it into data packets at the network and transport layers, and transmits it to the application layer. The application layer extracts the detected channel state information and utilizes it in joint source-channel coding. Uploading channel state information also requires a mechanism.

[0122] Channel state information can change dramatically over time, such as in high-speed mobile scenarios like autonomous driving. A channel state estimation module is designed. Based on a large amount of time-series channel state information training, this module is deployed at the application layer. This module predicts the channel state information at the next moment based on the previous channel state information. At the same time, the channel state estimation module can be further optimized using periodic actual feedback of channel state information.

[0123] Channel state information does not change dramatically over time, such as when a user watches a VR video indoors.

[0124] 1) Periodic upload: Upload channel status information once every certain period of time to complete the periodic transmission of channel status information.

[0125] 2) Change-driven upload: When the change range of certain parameters in the channel state information (SNR, multipath fading characteristics, etc.) exceeds a certain threshold, the channel state information is transmitted.

[0126] 3) Event-driven upload: Channel status information is uploaded when a specific event occurs, such as base station switching, network congestion, etc.

[0127] 4) Hybrid-driven upload: Combine the above mechanisms to jointly upload channel state information. For example, on the basis of periodic transmission, upload is performed when a drastic change in parameters in the channel state information is detected or a specific event is triggered.

[0128] The transmission and network layers of the transmitter 502 are used to group and vectorize the semantically encoded data based on the current transmission scenario, obtaining a semantic vector group set including at least one semantic vector group, where each semantic vector group corresponds one-to-one with a feature point group. Each semantic vector group in the semantic vector group set is segmented to obtain multiple segmented subvectors. Based on the importance level information, header information is added to each of the multiple segmented subvectors to obtain multiple data packets with importance levels. For a detailed description of the functions of the transmission and network layers of the transmitter 502, see step 104 of the illustrated embodiment of the information transmission method.

[0129] The data link layer of the transmitter 502 generates a checksum for each data packet corresponding to each segmentation subvector in the semantic vector group set based on the importance level information of the data packet. The checksum is then appended to the end of the data packet to obtain a transmission semantic frame for the data packet. For a detailed description of the functions of the data link layer of the transmitter 502, see step 105 of the illustrated embodiment of the information transmission method.

[0130] The physical layer of the transmitter 502 is used to modulate the transmission semantic frame to obtain and transmit the modulation information through the wireless channel 501. For a detailed description of the functions of the physical layer of the transmitter 502, please refer to step 106 of the embodiment of the information transmission method.

[0131] The physical layer of the receiving end 503 is used to demodulate the modulated information received from the wireless channel 501 to obtain a transmission semantic frame. For a detailed description of the functions of the physical layer of the receiving end 503, please refer to step 201 of the embodiment of the information processing method.

[0132] The data link layer of the receiving end 503 is used to verify the transmission semantics. For a detailed introduction to the functions of the data link layer of the receiving end 503, please refer to step 202 of the embodiment of the information processing method.

[0133] The transport and network layers of receiving end 503 are configured to, in response to the semantic frame passing the check, obtain a set of semantic vector groups based on the semantic frame; and obtain semantically encoded data based on the set of semantic vector groups. For a detailed description of the functions of the transport and network layers of receiving end 503, see steps 203 and 204 of the illustrated embodiment of the information processing method.

[0134] The application layer of the receiving end 503 performs source-channel joint decoding on the semantically coded data to obtain semantic features, and obtains the information to be transmitted based on the semantic features. For a detailed introduction to the functions of the application layer of the receiving end 503, please refer to steps 205 and 206 of the embodiment of the information processing method.

[0135] The semantic communication system provided by the embodiments of the present disclosure integrates semantic communication with traditional communication systems, transmits the semantic features of the information to be transmitted in blocks, and performs adaptive processing of the source and channel of the semantic features of the information to be transmitted. Specifically, the information to be transmitted is compressed at the semantic level through semantic extraction and semantic recovery, thereby reducing the amount of information to be transmitted; the source-channel joint coding can combine the channel information while retaining the correlation of the semantic features, and determine the most appropriate scheme between the source coding and the channel coding, thereby completing the optimal transmission of information with the least resource consumption; the importance of the semantic information is represented by the importance level information, so that different optional schemes are adopted during CRC verification to ensure error-free transmission of important information.

[0136] 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.

[0137] Figure 6 A schematic block diagram of an example electronic device 600 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.

[0138] like Figure 6 As shown, the device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. Various programs and data required for the operation of the device 600 can also be stored in the RAM 603. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0139] Various components in device 600 are connected to I / O interface 605, including an input unit 606, such as a keyboard, mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a magnetic disk, optical disk, etc.; and a communication unit 609, such as a network card, modem, wireless communication transceiver, etc. The communication unit 609 allows device 600 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0140] The computing unit 601 can be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 601 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 601 performs the various methods and processes described above, such as the information transmission method or the information processing method. For example, in some embodiments, the information transmission method or the information processing method can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as the storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed on the device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the method described above can be performed. Alternatively, in other embodiments, the computing unit 601 can be configured to perform the information transmission method or the information processing method by any other appropriate means (e.g., by means of firmware).

[0141] 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 that includes 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.

[0142] 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 information transmission device or information 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.

[0143] 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.

[0144] 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).

[0145] 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.

[0146] Computer systems may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The client and server relationship arises through computer programs running on the respective computers and having a client-server relationship to each other.

[0147] 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.

[0148] 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 method for transmitting information, the method comprising: Extracting semantic features of the information to be transmitted in the current transmission scenario; Based on the current transmission scenario and the semantic features, obtaining importance level information of a feature point group set including at least one feature point group; Performing source-channel joint coding on the semantic features to obtain semantic coding data; Based on the current transmission scenario, grouping and vectorizing the semantically coded data to obtain a semantic vector group set including at least one semantic vector group, wherein the semantic vector groups correspond one-to-one to the feature point groups; Obtaining a transmission semantic frame based on the importance level information and the semantic vector group set; The transmission semantic frame is modulated to obtain and transmit modulation information via a wireless channel.

2. The method according to claim 1, further comprising: An encryption algorithm is used to encrypt each semantic vector group in the semantic vector group set.

3. The method according to claim 1 or 2, wherein: The obtaining, based on the current transmission scenario and the semantic features, the importance level information of the feature point group set including at least one feature point group includes: Calculating the information entropy of each feature point in the semantic feature; Based on the current transmission scenario, the information entropy of the feature points is grouped to obtain a feature point group set of at least one feature point group; Calculating the normalized average value of the information entropy of each feature point group in the feature point group set; Based on the normalized average value, the importance level information of each feature point group in the feature point group set is obtained.

4. The method according to claim 3, wherein: The obtaining of the importance level information of each feature point group in the feature point group set based on the normalized average value includes: Based on the normalized average value of the information entropy of each feature point group in the feature point group set, a pre-set importance mapping set is queried to obtain the importance level information of each feature point group, and the importance mapping set is used to characterize the correspondence between the normalized average value and the importance level information.

5. The method according to claim 1, wherein The obtaining, based on the current transmission scenario and the semantic features, the importance level information of the feature point group set including at least one feature point group includes: Based on the current transmission scenario, the feature points in the semantic features are grouped to obtain a feature point group set including at least one feature point group; Based on the information entropy of the feature points in the semantic features, importance level information of each feature point group in the feature point group set is obtained.

6. The method according to claim 1 or 2, wherein: The grouping and vectorization processing of the semantically coded data based on the current transmission scenario to obtain a semantic vector group set including at least one semantic vector group includes: Based on a current transmission scenario, the semantically coded data is divided into semantically coded data groups belonging to high granularity and low granularity, respectively; Vectorization processing is performed on the semantically encoded data group to obtain a semantic vector group set including at least one semantic vector group.

7. The method according to claim 1 or 2, wherein: The obtaining of a transmission semantic frame based on the importance level information and the semantic vector group set includes: Segmenting each semantic vector group in the semantic vector group set to obtain a plurality of segmented subvectors; Based on the importance level information, adding header information to each of the plurality of slice subvectors to obtain a plurality of data packets having the importance level information; For the data packet corresponding to each slice subvector in the semantic vector group set, a check code of the data packet is generated based on the importance level information of the data packet, and the check code of the data packet is added to the end of the data packet to obtain a transmission semantic frame of the data packet.

8. The method according to claim 7, wherein: The step of generating a check code for a data packet corresponding to each slice subvector in the semantic vector group set based on the importance level information of the data packet includes any one of the following: generating, for a data packet corresponding to each slice subvector in the semantic vector group set, a first cyclic redundancy check code having a number of bits corresponding to the importance level information of the data packet based on the importance level information of the data packet; generating a first level check code based on the importance level information of the data packet, and using the first cyclic redundancy check code and the first level check code as a check code for the data packet; For the data packets corresponding to each segmentation subvector in the semantic vector group set, based on the importance level information of the data packet, a second cyclic redundancy check code having a check number corresponding to the importance level information of the data packet is generated; based on the importance level information of the data packet, a second level check code is generated, and the second cyclic redundancy check code and the second level check code are used as the check codes of the data packet.

9. An information processing method, the method comprising: Demodulate the received modulated information to obtain a transmission semantic frame; Verifying the transmission semantics; In response to the transmission semantic frame passing the check, obtaining a semantic vector group set based on the transmission semantic frame; Obtaining semantic coding data based on the semantic vector group set; performing source-channel joint decoding on the semantically coded data to obtain semantic features; Based on the semantic features, information to be transmitted is obtained.

10. The method according to claim 9, further comprising: Each semantic vector group in the semantic vector group set is decrypted.

11. An information transmission device, comprising: an extraction unit configured to extract semantic features of information to be transmitted in a current transmission scenario; a level obtaining unit configured to obtain importance level information of a feature point group set including at least one feature point group based on a current transmission scenario and the semantic feature; an encoding unit configured to perform source-channel joint encoding on the semantic feature to obtain semantically encoded data; a grouping unit configured to group and vectorize the semantically coded data based on a current transmission scenario to obtain a semantic vector group set including at least one semantic vector group, wherein the semantic vector groups correspond one-to-one to the feature point groups; a frame obtaining unit configured to obtain a transmission semantic frame based on the importance level information and the set of semantic vector groups; The modulation unit is configured to modulate the transmission semantic frame to obtain and transmit modulation information via a wireless channel.

12. An information processing device, comprising: a demodulation unit configured to demodulate the received modulated information to obtain a transmission semantic frame; A verification unit, configured to verify the transmission semantics; a semantic obtaining unit configured to obtain a set of semantic vector groups based on the transmission semantic frame in response to the transmission semantic frame passing the check; A coding obtaining unit configured to obtain semantic coding data based on the semantic vector group set; a decoding unit configured to perform source-channel joint decoding on the semantically coded data to obtain semantic features; The information obtaining unit is configured to obtain the information to be transmitted based on the semantic features.

13. A semantic communication system, comprising: The receiving and sending ends of information transmission through wireless channels; The application layer of the transmitting end is used to extract the semantic features of the information to be transmitted in the current transmission scenario; Based on the current transmission scenario and the semantic features, obtaining importance level information of a feature point group set including at least one feature point group; performing source-channel joint coding on the semantic features to obtain semantically coded data; wherein the channel state information used in the source-channel joint coding is obtained from the physical layer of the transmitting end; The transmission and network layer of the transmitting end is used to group and vectorize the semantically coded data based on the current transmission scenario to obtain a semantic vector group set including at least one semantic vector group, wherein the semantic vector group corresponds one-to-one with the feature point group; split each semantic vector group in the semantic vector group set to obtain a plurality of split subvectors; and add header information to each of the plurality of split subvectors based on the importance level information to obtain a plurality of data packets with importance levels; The data link layer of the transmitting end generates a check code for the data packet corresponding to each slice subvector in the semantic vector group set based on the importance level information of the data packet, and adds the check code of the data packet to the end of the data packet to obtain a transmission semantic frame of the data packet; The physical layer of the transmitting end is used to modulate the transmission semantic frame to obtain and transmit modulation information through the wireless channel; The physical layer of the receiving end is used to demodulate the modulated information received from the wireless channel to obtain a transmission semantic frame; The data link layer of the receiving end is used to verify the transmission semantics; The transmission and network layer of the receiving end is used to obtain a semantic vector group set based on the transmission semantic frame in response to the transmission semantic frame passing the check; and obtain semantic coding data based on the semantic vector group set; The application layer of the receiving end performs source-channel joint decoding on the semantically coded data to obtain semantic features; and obtains information to be transmitted based on the semantic features.

14. An electronic device, characterized in that: include: 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 according to any one of claims 1 to 10.

15. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 10.

Images