Distributed semantic communication authentication method and device, electronic equipment and storage medium

Abstract

The invention provides a distributed semantic communication authentication method and device, electronic equipment and a storage medium, and relates to the technical field of computers, in particular to the technical field of semantic communication. According to the specific implementation scheme, the method comprises the following steps: receiving a channel semantic fingerprint sent by at least one distributed node to obtain a first channel fingerprint; receiving a channel fingerprint sent by the sending end to obtain a second channel fingerprint; constructing a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint; and authenticating the legality of the transmission channel by using a pre-trained identity authentication model according to the multi-dimensional channel fingerprint matrix to obtain a legality authentication result. The authentication accuracy and robustness are improved, the communication safety and reliability are ensured, and the authentication efficiency and the response speed are improved through automatic processing of the deep learning model.

Description

Technical Field

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

[0002] With the rapid development of wireless communication technology, especially in the expected deployment of 6G networks, the number of radio devices around the world has increased dramatically to achieve wider network coverage and richer application scenarios. However, the openness of wireless transmission media also brings security risks. Potential attackers can exploit this openness to forge or manipulate identity information, thereby impersonating authorized users and accessing sensitive data, which may even cause significant economic losses or security threats.

[0003] In existing wireless communication systems, terminal authentication mainly relies on encryption mechanisms implemented by the core network, such as EAP-AKA (Extensible Authentication Protocol-Authentication and Key Agreement). Although these mechanisms provide security to a certain extent, they face limitations in security, complexity and compatibility in a massive device environment. In addition, physical layer authentication (PLA), as a supplementary means, can provide natural protection and personalized identification for devices by utilizing unique characteristics derived from communication links and location-specific attributes, but its authentication accuracy and robustness in high-density scenarios still need to be improved. Summary of the invention

[0004] The present disclosure provides a distributed semantic communication authentication method, device, electronic device and storage medium.

[0005] According to one aspect of the present disclosure, a distributed semantic communication authentication method is provided, the method comprising:

[0006] Receiving a channel semantic fingerprint sent from at least one distributed node to obtain a first channel fingerprint;

[0007] Receiving a channel fingerprint sent from a sending end to obtain a second channel fingerprint;

[0008] Constructing a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint;

[0009] The legitimacy of the transmission channel is authenticated using a pre-trained identity authentication model according to the multi-dimensional channel fingerprint matrix to obtain a legitimacy authentication result.

[0010] According to another aspect of the present disclosure, a distributed semantic communication authentication device is provided, comprising:

[0011] A first receiving module, configured to receive a channel semantic fingerprint sent from at least one distributed node to obtain a first channel fingerprint;

[0012] A second receiving module is used to receive the channel fingerprint sent from the sending end to obtain a second channel fingerprint;

[0013] A construction module, configured to construct a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint;

[0014] The authentication module is used to authenticate the legitimacy of the transmission channel using a pre-trained identity authentication model according to the multi-dimensional channel fingerprint matrix to obtain a legitimacy authentication result.

[0015] According to a third aspect of the present disclosure, there is provided an electronic device, including:

[0016] at least one processor; and

[0017] a memory communicatively connected to the at least one processor; wherein,

[0018] 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 so that the at least one processor can execute any method in any of the above technical solutions.

[0019] According to a fourth 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 any one of the methods described in the above technical solutions.

[0020] According to a fifth aspect of the present disclosure, a computer program product is provided, including a computer program, wherein the computer program implements any one of the methods described in the above technical solutions when executed by a processor.

[0021] The present disclosure provides a distributed semantic communication authentication method, apparatus, device and storage medium. The present disclosure receives a channel semantic fingerprint from at least one distributed node to obtain a first channel fingerprint, and directly obtains the channel fingerprint from the transmitting end to obtain a second channel fingerprint, and constructs a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint. The matrix integrates rich channel information from different sources. This fusion not only enhances the comprehensiveness of the channel characteristics, but also improves the robustness of the authentication process. Furthermore, the multi-dimensional channel fingerprint matrix is ​​analyzed using a pre-trained identity authentication model, so that the legitimacy of the transmission channel can be effectively authenticated and a legitimacy authentication result can be obtained. This process not only improves the accuracy and robustness of the authentication, ensures the security and reliability of the communication, but also improves the authentication efficiency and response speed through the automated processing of the deep learning model. Therefore, the method achieves a more efficient, more accurate and more secure physical layer authentication effect, providing a strong security guarantee for wireless communications.

[0022] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0024] Figure 1 is a schematic diagram of the steps of the distributed semantic communication authentication method in an embodiment of the present disclosure;

[0025] Figure 2 is a flowchart corresponding to a distributed semantic communication authentication method in an embodiment of the present disclosure;

[0026] Figure 3 is a scene graph corresponding to the distributed semantic communication authentication method in an embodiment of the present disclosure;

[0027] Figure 4 A principle block diagram of a distributed semantic communication authentication device in an embodiment of the present disclosure;

[0028] Figure 5 It is a block diagram of an electronic device used to implement the distributed semantic communication authentication method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0029] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art 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.

[0030] In the related technical solutions, although a variety of physical layer authentication (PLA) methods have been proposed to enhance the security of wireless communication systems, there are still some technical problems and challenges. For example, the first authentication method is a threshold-free multi-attribute physical layer authentication method, which obtains the attribute value of the physical layer authentication fingerprint of the transmission channel according to the received signal data by utilizing multi-antenna technology. This method combines the multiple attributes of the physical layer authentication fingerprint, provides multi-dimensional protection for the wireless communication system, effectively reduces the risk of authentication failure due to attackers imitating certain attributes, and reduces the probability of authentication errors. However, this method may face challenges in high-density device scenarios, because the differences in signal strength indication (RSSI), carrier frequency offset (CFO) and channel impulse response (CIR) between devices with similar distances may not be large, which makes it difficult for physical layer verification based on these features to effectively distinguish different devices.

[0031] The second authentication method is a physical layer authentication method based on multi-time slot channel characteristics. This method performs physical layer authentication of the sender and receiver in each time slot based on the physical layer characteristic information of the previous time slot and the status information of the current time slot, and updates the physical layer characteristic information of the current time slot. This method can achieve high security and reliability of data transmission. However, in dynamic scenarios, due to large channel changes, this solution may misjudge the identification of user identities, such as misjudging legitimate users as unauthorized users, which may affect the system availability and user experience.

[0032] In summary, although the existing technology has improved security to a certain extent when providing physical layer authentication, it still needs to be improved in terms of differentiation ability in high-density device environments, stability and accuracy in dynamic scenarios. Therefore, a more accurate and robust physical layer authentication method is urgently needed.

[0033] In order to solve the above technical problems, the present disclosure provides a distributed semantic communication authentication method, see Figure 1 As shown, Figure 1 : is a schematic diagram of the steps of a distributed semantic communication authentication method in an embodiment of the present disclosure, the method is applied to a receiving end, and the method includes:

[0034] Step S101: receiving a channel semantic fingerprint sent from at least one distributed node to obtain a first channel fingerprint.

[0035] Among them, distributed nodes (also referred to as distributed collaboration points) refer to multiple collaborative units or devices that are geographically dispersed in a wireless communication network. These nodes are usually equipped with wireless communication capabilities, can independently collect and process channel state information (CSI), and participate in the physical layer authentication (PLA) process. In the embodiment of the present disclosure, the number of distributed nodes is at least one, which can be one or two, three or more. The "first channel fingerprint" refers to the channel semantic fingerprint received from at least one distributed node, which is a semantic information extracted by deep learning technology that can represent the unique channel characteristics of each distributed node.

[0036] Specifically, the specific implementation process of this embodiment includes that each distributed node first collects channel state information (CSI) related to the communication link, which contains the physical characteristics of the signal during propagation, such as signal strength, delay, frequency offset, etc. Then, using the channel fingerprint semantic image model deployed on the node, these CSI data are converted into a channel fingerprint image rich in semantic features, namely the first channel fingerprint. This process not only captures the unique physical properties of signal propagation, but also makes these features more prominent and easy to process through the abstraction and compression of the deep learning model. Finally, these first channel fingerprints are used to construct a multi-dimensional channel fingerprint matrix to provide key input data for the subsequent identity authentication model, thereby realizing accurate authentication of the legitimacy of the communication link. This method is not only conducive to improving the robustness and accuracy of authentication, but also enhances the defense capability of the entire communication system against potential attacks by leveraging the collaboration of distributed nodes.

[0037] Step S102: receiving a channel fingerprint sent from a transmitting end to obtain a second channel fingerprint.

[0038] Specifically, the "second channel fingerprint" refers to the channel fingerprint directly obtained from the transmitting end (e.g., user equipment or base station), which represents the physical characteristics of the communication link between the transmitting end and the receiving end (such as Bob). The specific implementation process of this embodiment includes that the transmitting end sends specific signals or pilots through its wireless interface. These signals are affected by various factors during propagation, including path loss, multipath effect, signal attenuation, etc., thus forming unique channel state information (CSI). The receiving end (Bob) uses its hardware device to capture these signals and perform digital processing to extract CSI data. These data are then converted into channel fingerprints, that is, the second channel fingerprints, which are direct reflections of the signal propagation characteristics and contain key information such as signal strength, arrival time, phase offset, etc. The second channel fingerprint provides direct channel information that is not affected by a third party for constructing a multi-dimensional channel fingerprint matrix and is a key step in realizing physical layer authentication and ensuring communication security.

[0039] Step S103: Construct a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint.

[0040] Specifically, the "multi-dimensional channel fingerprint matrix" refers to integrating channel fingerprint information from different sources into a unified data structure for in-depth analysis and processing. Specifically, the "first channel fingerprint" usually refers to the channel semantic fingerprint collected from distributed nodes, which is extracted from the channel state information (CSI) through a deep learning model and reflects the propagation characteristics of the signal at different nodes; while the "second channel fingerprint" is the channel fingerprint directly obtained from the transmitting end, which provides direct information about the channel from the transmitting end to the receiving end.

[0041] After obtaining the first channel fingerprint and the second channel fingerprint, they are organized according to certain rules and formats to form a multi-dimensional matrix. This matrix not only contains the distribution information of the signal in time and space but also may contain multi-dimensional data such as signal strength, phase, and frequency, thus providing a comprehensive and three-dimensional view of the channel characteristics for the subsequent authentication model. Through this multi-dimensional integration, this solution can more accurately capture and distinguish the unique characteristics in the communication link, enhance the accuracy and robustness of identity authentication, and at the same time improve the detection ability for potential attacks. The construction of this multi-dimensional channel fingerprint matrix is the basis for realizing efficient and reliable physical layer authentication and provides strong support for wireless communication security.

[0042] Step S104: Authenticate the legality of the transmission channel according to the multi-dimensional channel fingerprint matrix using a pre-trained authentication model to obtain a legality authentication result.

[0043] Specifically, after obtaining the multi-dimensional channel fingerprint matrix, since this matrix provides a comprehensive data view for the deep learning model, the model can capture the unique characteristics of the communication link.

[0044] In addition, the "pre-trained authentication model" refers to a model that is pre-trained by a machine learning algorithm using historical data and labels, which can identify and distinguish between legitimate and illegitimate channel fingerprint patterns. After obtaining the multi-dimensional channel fingerprint matrix and the authentication model, the authentication model is used to process the multi-dimensional channel fingerprint matrix. By analyzing the complex data in the matrix, the authentication model can learn the pattern of legitimate channel fingerprints and judge the legitimacy of the current transmission channel accordingly. It should be noted that the output of the authentication model is a legitimacy authentication result, that is, a predicted label indicating whether the channel is used by a legitimate user. This process not only improves the accuracy of authentication, but also enhances the system's defense capabilities against potential attacks, ensuring the security and reliability of wireless communications. Through this advanced authentication mechanism, this solution can effectively identify legitimate users and prevent unauthorized access, thereby protecting the communication link from malicious attacks.

[0045] The present disclosure provides a distributed semantic communication authentication method, apparatus, device and storage medium. The present disclosure receives a channel semantic fingerprint from at least one distributed node to obtain a first channel fingerprint, and directly obtains the channel fingerprint from the transmitting end to obtain a second channel fingerprint, and constructs a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint. The matrix integrates rich channel information from different sources. This fusion not only enhances the comprehensiveness of the channel characteristics, but also improves the robustness of the authentication process. Furthermore, the multi-dimensional channel fingerprint matrix is ​​analyzed using a pre-trained identity authentication model, so that the legitimacy of the transmission channel can be effectively authenticated and a legitimacy authentication result can be obtained. This process not only improves the accuracy and robustness of the authentication, ensures the security and reliability of the communication, but also improves the authentication efficiency and response speed through the automated processing of the deep learning model. Therefore, the method achieves a more efficient, more accurate and more secure physical layer authentication effect, providing a strong security guarantee for wireless communications.

[0046] In some optional embodiments, receiving a channel semantic fingerprint sent from at least one distributed node to obtain a first channel fingerprint includes:

[0047] receiving a channel semantic fingerprint sent from at least one distributed node, where the channel semantic fingerprint is semantically compressed channel state information;

[0048] A pre-trained semantic model is used to perform semantic recovery on the received channel semantic fingerprint to obtain a first channel fingerprint.

[0049] Specifically, "channel semantic fingerprint" refers to the semantically compressed channel state information (CSI), which is a key information extracted by a deep learning model that can represent the signal propagation characteristics. It converts the complex features in the original CSI data into a semantic form that is easier to process and transmit. Furthermore, when the distributed nodes collect CSI data, the data will be input into a pre-trained semantic model, which can identify and compress the core features in the CSI data to generate channel semantic fingerprints. Subsequently, these semantically compressed channel semantic fingerprints are sent to the receiving end. At the receiving end, the same pre-trained semantic model is used to perform semantic recovery on the received channel semantic fingerprints, that is, the semantic model will reconstruct the semantic representation of the original CSI data based on the compressed features it has learned, thereby obtaining the "first channel fingerprint". This process not only reduces the amount of data required for transmission and improves transmission efficiency, but also ensures that the key features of the channel can be accurately captured and recovered even when the channel conditions change through the semantic recovery capability of the model, providing a solid foundation for subsequent identity authentication and communication security.

[0050] In this way, by receiving channel semantic fingerprints from at least one distributed node, these channel semantic fingerprints are used as semantically compressed channel state information, thereby effectively reducing the volume of transmitted data and reducing the communication burden. Subsequently, the received channel semantic fingerprints are semantically restored using a pre-trained semantic model to obtain the first channel fingerprint. This process not only retains the core features of the CSI data, but also improves the efficiency and accuracy of signal processing. This method of combining semantic compression and recovery enhances the adaptability to channel changes, improves the accuracy of identity authentication, and ensures efficient and reliable data transmission in a changing communication environment, thereby providing wireless communication systems with greater security and robustness.

[0051] In some optional embodiments, before receiving the channel semantic fingerprint sent from at least one distributed node, the method further includes:

[0052] Obtain channel state sample data;

[0053] The first model to be trained is trained according to the channel state sample data to obtain a semantic model.

[0054] Specifically, "channel state sample data" refers to a series of channel state information (CSI) samples collected from wireless communication links. These samples contain various physical characteristics of the signal during propagation, such as signal strength, arrival time, phase information, etc. They are a direct reflection of the characteristics of the wireless channel. "Semantic model" refers to a deep learning model that can extract the semantic features of the channel from the CSI sample data, that is, the key information that can represent the channel propagation characteristics.

[0055] The training process of the semantic model includes collecting channel state sample data from multiple communication links, which form the basis for training the semantic model. Then, the first model to be trained is trained using these sample data. During the training process, the model gradually adjusts its internal parameters by learning the patterns and associations in the sample data to identify and extract the semantic features in the channel state information. Ultimately, the trained semantic model can accurately extract these semantic features from new CSI samples, providing support for subsequent channel fingerprint construction and identity authentication. This process not only improves the processing efficiency of channel state information, but also enhances the model's adaptability and recognition capabilities to channel changes, thereby providing strong protection for the identity authentication and security of wireless communication systems.

[0056] In this way, by obtaining channel state sample data and using these data to train the first model to be trained, an accurate semantic model can be constructed. This semantic model can deeply understand and extract complex features in channel state information and convert it into useful semantic information. Such processing not only improves the readability and usability of channel data, but also enhances the model's ability to recognize channel changes, thereby achieving more efficient and accurate physical layer authentication in wireless communications. This data-driven approach enables the system to adaptively respond to dynamic changes in the channel, improves the robustness of communication, and reduces the complexity of system design by reducing reliance on artificial feature engineering, ultimately improving the performance and security of the entire communication system.

[0057] In some optional embodiments, training the first to-be-trained model according to the channel state sample data to obtain a semantic model includes:

[0058] Extract features from channel state sample data to obtain key features;

[0059] Convert the key features into image format to obtain a two-dimensional signal fingerprint image set;

[0060] The first model to be trained is trained according to the two-dimensional signal fingerprint image set until the model converges to obtain a semantic model.

[0061] Specifically, after obtaining the channel state sample data, the channel state sample data is firstly subjected to feature extraction. This step involves identifying and extracting key information that can represent the signal propagation characteristics from the original channel state information (CSI), such as the amplitude, phase, time delay, etc. of the signal. This information can reveal the unique mode of signal propagation in the wireless channel. Then, these key features are converted into image format to form a two-dimensional signal fingerprint image set. This process makes it possible to visualize the channel features and further deep learning processing, and also facilitates the use of image processing technology to enhance the recognition ability of features. Finally, the first model to be trained is trained using these two-dimensional signal fingerprint image sets until the model converges, thereby obtaining a deep learning model that can understand and identify channel features. This semantic model can learn deep patterns and associations from the channel fingerprint image, providing strong support for identity authentication and security in wireless communication systems. Through such an implementation process, not only the processing efficiency and accuracy of channel data are improved, but also the system's adaptability to channel changes is enhanced, laying the foundation for efficient and reliable physical layer authentication.

[0062] In this way, by extracting features from the channel state sample data to obtain key features, and then converting these features into a two-dimensional signal fingerprint image set, it is possible to effectively represent complex channel information in an intuitive image form, which not only facilitates the processing of deep learning models, but also enhances the identifiability and comparability of features. Furthermore, the first model to be trained is trained using these two-dimensional signal fingerprint image sets until the model converges, and the resulting semantic model can accurately capture and understand the semantic information of the channel, thereby significantly improving the accuracy and robustness of physical layer authentication in wireless communication systems. This semantic model training method based on deep learning enables the system to automatically identify and distinguish between legal and illegal communication links, providing a more solid foundation for wireless communication security, and also providing the possibility of efficient and intelligent channel management and optimization.

[0063] In some optional embodiments, constructing a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint includes:

[0064] The first channel fingerprint and the second channel fingerprint are combined to obtain a multi-dimensional channel fingerprint matrix.

[0065] Specifically, the "first channel fingerprint" and "second channel fingerprint" refer to the semantically processed channel state information obtained from the distributed nodes and the transmitter, respectively. They each capture the unique physical characteristics of the signal propagating in different links and are converted into information-rich semantic representations. After obtaining the first channel fingerprint and the second channel fingerprint, the two channel fingerprint data sets are integrated together through a carefully designed merging process to form a multi-dimensional channel fingerprint matrix. This matrix not only contains the spatial distribution information of the signal, but may also contain data in multiple dimensions such as time and frequency, thus providing a comprehensive channel feature view for the model. The construction of the multi-dimensional channel fingerprint matrix enables this solution to use deep learning models to analyze channel characteristics from different angles and levels, greatly enhancing the system's ability to identify the legitimacy of the communication link, improving the accuracy and robustness of identity authentication, and providing a more solid foundation for wireless communication security.

[0066] In this way, by merging the first channel fingerprint and the second channel fingerprint to obtain a multi-dimensional channel fingerprint matrix, this process significantly improves the authentication capability and security of the wireless communication system. Since the merged matrix integrates channel information from different sources, it provides a comprehensive view of channel characteristics, allowing the system to capture and analyze the characteristics of signal propagation from multiple dimensions. This multi-dimensional analysis enhances the ability to perceive subtle changes in signals and improves the accuracy of distinguishing legitimate users from potential attackers. In addition, the multi-dimensional channel fingerprint matrix also increases the difficulty for attackers to simulate or crack channel characteristics, thereby providing a more solid line of defense for communication security at the physical layer, ensuring the legitimacy of the communication link and the security of data transmission.

[0067] In some optional embodiments, the legitimacy of the transmission channel is authenticated using a pre-trained identity authentication model according to the multi-dimensional channel fingerprint matrix to obtain a legitimacy authentication result, including:

[0068] The multi-dimensional channel fingerprint matrix is ​​input into the pre-trained identity authentication model, and the legitimacy of the transmission channel is authenticated by the identity authentication model to obtain the legitimacy authentication result.

[0069] Specifically, the "multi-dimensional channel fingerprint matrix" is a data structure that integrates rich information from different channel paths. It contains multiple physical layer characteristics of the signal, such as signal strength, time delay, frequency offset, etc. These characteristics together constitute a comprehensive channel description. The "pre-trained authentication model" refers to a model trained by a machine learning algorithm using historical data and labels. This model can identify and distinguish between legitimate and illegitimate channel fingerprint patterns.

[0070] In the specific implementation process of this embodiment, the multi-dimensional channel fingerprint matrix is ​​first provided as input data to the identity authentication model. The identity authentication model analyzes the input channel fingerprint matrix through its learned complex algorithms and weights, identifies the channel characteristic patterns of legitimate users, and compares them with known legitimate channel fingerprints. Through this comparison, the identity authentication model can determine the legitimacy of the current transmission channel and finally output a legitimacy authentication result, that is, a predicted label indicating whether the channel is used by a legitimate user. This process not only improves the accuracy of authentication, but also enhances the system's defense capabilities against potential attacks, ensuring the security and reliability of wireless communications. Through this advanced authentication mechanism, the present solution can effectively identify legitimate users and prevent unauthorized access, thereby protecting the communication link from malicious attacks.

[0071] In this way, by inputting the multi-dimensional channel fingerprint matrix into the pre-trained authentication model, the learning and prediction capabilities of the model can be fully utilized to accurately authenticate the legitimacy of the transmission channel. This authentication method not only improves the automation and efficiency of the authentication process, but also because the authentication model is trained on historical data, it can identify complex channel feature patterns, thereby effectively distinguishing legitimate users from potential attackers. Such a mechanism significantly enhances the security of wireless communication systems, reduces the risk of unauthorized access, and ensures the confidentiality and integrity of data transmission. In addition, the accuracy and reliability of the legitimacy authentication results have also been improved, providing strong technical support for maintaining the stability and trust of communication networks.

[0072] In some optional embodiments, the method further comprises:

[0073] Receiving a channel semantic fingerprint sent by at least one distributed node, and obtaining a distributed node channel semantic fingerprint;

[0074] Receive the channel fingerprint sent by the sender and obtain the original channel fingerprint;

[0075] The second model to be trained is trained according to the distributed node channel fingerprint and the original channel fingerprint until the model converges to obtain an identity authentication model.

[0076] Specifically, "distributed node channel semantic fingerprint" refers to the semantically processed channel state information received from at least one distributed node. This information extracts key features from the original channel data through a deep learning model and converts it into a semantic form for easy analysis and processing. "Raw channel fingerprint" refers to the channel state information obtained directly from the sender without semantic processing, which is a direct reflection of the signal propagation characteristics.

[0077] The training process of the authentication model includes: first, collecting the channel semantic fingerprints from the distributed nodes and the original channel fingerprints of the sender, these two types of information together constitute the training data set. Then, use these data to train the second model to be trained. During the training process, the model learns how to identify the characteristics of legitimate users and potential attackers from these channel fingerprints. As the training progresses, the model continuously adjusts its internal parameters until it reaches a convergence state, at which time the model can accurately classify new channel fingerprint data, thereby obtaining a fully trained authentication model. This authentication model can authenticate the legitimacy of the transmission channel in wireless communication, improving the security of the communication system and the accuracy of authentication. Through this training method that combines distributed node and sender channel fingerprints, this scheme can fully utilize the advantages of multi-source data, enhance the generalization ability and robustness of the model, and provide a more solid foundation for wireless communication security.

[0078] In this way, by receiving the channel semantic fingerprint sent by at least one distributed node to obtain the distributed node channel semantic fingerprint, and receiving the original channel fingerprint from the sender, this scheme can comprehensively utilize the channel information from different sources, which enhances the comprehensiveness and diversity of the channel data. Furthermore, these rich channel fingerprint data are used to train the second model to be trained until the model converges, and an accurate identity authentication model is obtained. This model can accurately identify and distinguish legitimate users from potential attackers, significantly improving the accuracy and robustness of physical layer authentication in wireless communication systems. In addition, this training method that combines distributed node and sender channel fingerprints enables the identity authentication model to have stronger generalization capabilities and can adapt to changing communication environments, thereby ensuring communication security while also improving the reliability and efficiency of the system.

[0079] In some optional embodiments, see Figure 2 , Figure 2 This is a flowchart corresponding to the distributed semantic communication authentication method in an embodiment of the present disclosure. The flowchart shows the implementation process of a collaborative physical layer authentication method enabled by distributed semantic communication, involving the collaborative work of multiple distributed collaboration points and the receiving end Bob. The following is a detailed description of each step in the flowchart:

[0080] (1) CSI training set: First, a training data set of channel state information (CSI) is collected. This data will be used to train the channel fingerprint semantic model (referred to as the semantic model for short).

[0081] (2) Extract channel fingerprints: Channel fingerprints are extracted from the CSI training set. These fingerprints are physical layer representations of the characteristics of wireless communication links.

[0082] (3) Training a channel fingerprint semantic model: Use the extracted channel fingerprint to train a channel fingerprint semantic model that can understand and recognize the semantic features of the channel fingerprint.

[0083] (4) Model deployment: The trained channel fingerprint semantic model is deployed to each distributed collaboration point (C1, C2, C3) and the receiving end Bob.

[0084] (5) Distributed collaboration points C1, C2, and C3: These collaboration points are responsible for collecting and processing CSI data from different geographical locations and sending the processed channel fingerprints to the receiving end Bob through semantic transmission.

[0085] (6) Alice or Eve’s CSI acquisition: The receiving end Bob obtains CSI data from Alice (legitimate user) or Eve (potential attacker).

[0086] (7) Input node and Bob: As an input node, Bob receives channel fingerprint data from the distributed cooperation points and compares it with the CSI data obtained by himself.

[0087] (8) Multi-dimensional fingerprint authentication: Bob uses the deployed channel fingerprint semantic model to process the collected CSI data, construct a multi-dimensional channel fingerprint matrix, and perform identity authentication.

[0088] (9) Distinguishing between Alice and Eve: Through the identity authentication model, Bob can determine whether the other party in the current communication is Alice or Eve, thereby ensuring the security of the communication.

[0089] (10) Authentication: Ultimately, Bob decides whether to continue communicating with the other party based on the results of identity authentication, thus achieving physical layer security authentication.

[0090] This flowchart clearly shows the implementation steps of the distributed collaborative physical layer authentication method, from channel fingerprint extraction and model training to model deployment and actual identity authentication process. Through this distributed collaboration and application of semantic models, the solution can effectively improve the accuracy and robustness of physical layer authentication and provide stronger security for wireless communication systems.

[0091] See also Figure 3 , Figure 3It is a scene diagram corresponding to the distributed semantic communication authentication method in an embodiment of the present disclosure. The scene diagram shows a physical layer authentication process, which includes a legitimate user Alice, a potential attacker Eve, and multiple distributed collaboration points (C1, C2, ..., CK-1, CK) and a receiving end Bob. Different line types are used in the figure to distinguish different types of links: dotted arrows represent spoofing links, solid arrows represent legitimate links, and bold arrows represent fingerprint transmission links. The following is a detailed description of each step in the figure:

[0092] Fingerprint Estimation: Alice and Eve communicate with distributed coordination points (C1, C2, ..., CK) to estimate channel fingerprints. This step corresponds to ① in the figure, where the link between Alice and the coordination point is legitimate (solid arrow), while the link between Eve and the coordination point is spoofed (dashed arrow).

[0093] Fingerprint Transmission: The distributed cooperation point transmits the estimated channel fingerprint to the receiving end Bob. This step corresponds to ② in the figure, represented by bold arrows. These links are used to transmit channel fingerprint information.

[0094] Fingerprint Aggression and Authentication: Bob receives the channel fingerprint from the distributed collaboration point and compares and aggregates it with the channel fingerprint obtained directly from Alice or Eve. This step corresponds to ③ in the figure. Bob uses the pre-trained authentication model to determine the legitimacy of the communication by analyzing these multi-dimensional channel fingerprint matrices.

[0095] In the figure, Eve tries to imitate Alice's channel fingerprint by spoofing the link, but Bob can analyze the channel characteristics from multiple dimensions by receiving channel fingerprints from multiple collaboration points, thereby identifying Eve's deception. The distinction between legitimate links and spoofed links, as well as the security of fingerprint transmission, are the key to achieving physical layer authentication. This scenario diagram clearly shows the implementation steps of the distributed collaborative physical layer authentication method, from the estimation and transmission of channel fingerprints to the final aggregation and authentication process. Through this distributed collaboration and multi-dimensional channel fingerprint analysis, the scheme can effectively improve the accuracy and robustness of physical layer authentication, providing stronger security for wireless communication systems.

[0096] The following describes an apparatus embodiment of the present application, which can be used to execute the distributed semantic communication authentication method in the above-mentioned embodiment of the present application. For details not disclosed in the apparatus embodiment of the present application, please refer to the above-mentioned embodiment of the distributed semantic communication authentication method of the present application.

[0097] The present disclosure also provides a distributed semantic communication authentication device 400, such as Figure 4 As shown, the device comprises:

[0098] A first receiving module 401 is configured to receive a channel semantic fingerprint sent from at least one distributed node to obtain a first channel fingerprint;

[0099] The second receiving module 402 is used to receive the channel fingerprint sent from the sending end to obtain a second channel fingerprint;

[0100] A construction module 403 is used to construct a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint;

[0101] The authentication module 404 is used to authenticate the legitimacy of the transmission channel using a pre-trained identity authentication model according to the multi-dimensional channel fingerprint matrix to obtain a legitimacy authentication result.

[0102] In some optional embodiments, the first receiving module 401 receives a channel semantic fingerprint sent from at least one distributed node to obtain a first channel fingerprint, including:

[0103] receiving a channel semantic fingerprint sent from at least one distributed node, where the channel semantic fingerprint is semantically compressed channel state information;

[0104] A pre-trained semantic model is used to perform semantic recovery on the received channel semantic fingerprint to obtain a first channel fingerprint.

[0105] In some optional embodiments, before receiving the channel semantic fingerprint sent from at least one distributed node, the first receiving module 401 is further used to obtain channel state sample data;

[0106] The first model to be trained is trained according to the channel state sample data to obtain a semantic model.

[0107] In some optional embodiments, the first receiving module 401 trains the first to-be-trained model according to the channel state sample data to obtain a semantic model, including:

[0108] Extract features from channel state sample data to obtain key features;

[0109] Convert the key features into image format to obtain a two-dimensional signal fingerprint image set;

[0110] The first model to be trained is trained according to the two-dimensional signal fingerprint image set until the model converges to obtain a semantic model.

[0111] In some optional embodiments, the construction module 403 constructs a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint, including:

[0112] The first channel fingerprint and the second channel fingerprint are combined to obtain a multi-dimensional channel fingerprint matrix.

[0113] In some optional embodiments, the authentication module 404 authenticates the legitimacy of the transmission channel using a pre-trained identity authentication model according to the multi-dimensional channel fingerprint matrix to obtain a legitimacy authentication result, including:

[0114] The multi-dimensional channel fingerprint matrix is ​​input into the pre-trained identity authentication model, and the legitimacy of the transmission channel is authenticated by the identity authentication model to obtain the legitimacy authentication result.

[0115] In some optional embodiments, the apparatus further includes a training module, configured to receive a channel semantic fingerprint sent by at least one distributed node, and obtain a distributed node channel semantic fingerprint;

[0116] Receive the channel fingerprint sent by the sender and obtain the original channel fingerprint;

[0117] The second model to be trained is trained according to the distributed node channel fingerprint and the original channel fingerprint until the model converges to obtain an identity authentication model.

[0118] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0119] 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.

[0120] Figure 5 A schematic block diagram of an example electronic device 500 that can be used to implement an embodiment 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 processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0121] like Figure 5 As shown, the electronic device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0122] A number of components in the device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0123] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 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 running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as a distributed semantic communication authentication method. For example, in some embodiments, the distributed semantic communication authentication method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 500 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by the computing unit 501, one or more steps of the applet distribution described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to perform the distributed semantic communication authentication method in any other appropriate manner (e.g., by means of firmware).

[0124] The various embodiments of the systems and techniques described above in this specification can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex 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 interpretable on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that receives data and instructions from, and transmits data and instructions to, a storage system, at least one input device, and at least one output device.

[0125] The program code for implementing the methods 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, special purpose computer, or other programmable data processing device such that the program codes, when executed by the processor or controller, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.

[0126] 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 connection with an instruction execution system, apparatus, or device. 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, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0127] 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).

[0128] The systems and techniques described herein may 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 with 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 may 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.

[0129] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0130] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0131] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A distributed semantic communication authentication method, the method comprising: Receiving a channel semantic fingerprint sent from at least one distributed node to obtain a first channel fingerprint; Receiving a channel fingerprint sent from a sending end to obtain a second channel fingerprint; Constructing a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint; The legitimacy of the transmission channel is authenticated using a pre-trained identity authentication model according to the multi-dimensional channel fingerprint matrix to obtain a legitimacy authentication result.

2. The method according to claim 1, wherein: The receiving a channel semantic fingerprint sent from at least one distributed node to obtain a first channel fingerprint includes: Receiving a channel semantic fingerprint sent from at least one distributed node, where the channel semantic fingerprint is semantically compressed channel state information; A pre-trained semantic model is used to perform semantic recovery on the received channel semantic fingerprint to obtain the first channel fingerprint.

3. The method according to claim 1, wherein: Before receiving the channel semantic fingerprint sent from at least one distributed node, the method further includes: Obtain channel state sample data; The first model to be trained is trained according to the channel state sample data to obtain the semantic model.

4. The method according to claim 3, wherein: The step of training the first model to be trained according to the channel state sample data to obtain the semantic model includes: Extracting features from the channel state sample data to obtain key features; Converting the key features into an image format to obtain a two-dimensional signal fingerprint image set; The first model to be trained is trained according to the two-dimensional signal fingerprint image set until the model converges to obtain the semantic model.

5. The method according to claim 1, wherein: The constructing a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint includes: The first channel fingerprint and the second channel fingerprint are combined to obtain the multi-dimensional channel fingerprint matrix.

6. The method according to claim 1, wherein: The method of authenticating the legitimacy of the transmission channel using a pre-trained identity authentication model according to the multi-dimensional channel fingerprint matrix to obtain a legitimacy authentication result includes: The multi-dimensional channel fingerprint matrix is ​​input into a pre-trained identity authentication model, and the legitimacy of the transmission channel is authenticated through the identity authentication model to obtain a legitimacy authentication result.

7. The method according to any one of claims 1 to 6, further comprising: Receiving a channel semantic fingerprint sent by at least one distributed node, and obtaining a distributed node channel semantic fingerprint; Receive the channel fingerprint sent by the sender and obtain the original channel fingerprint; The second model to be trained is trained according to the distributed node channel fingerprint and the original channel fingerprint until the model converges to obtain an identity authentication model.

8. A distributed semantic communication authentication device, the device comprising: A first receiving module, configured to receive a channel semantic fingerprint sent from at least one distributed node to obtain a first channel fingerprint; A second receiving module is used to receive the channel fingerprint sent from the sending end to obtain a second channel fingerprint; A construction module, configured to construct a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint; The authentication module is used to authenticate the legitimacy of the transmission channel using a pre-trained identity authentication model according to the multi-dimensional channel fingerprint matrix to obtain a legitimacy authentication result.

9. The device according to claim 8, wherein: The first receiving module receives a channel semantic fingerprint sent from at least one distributed node to obtain a first channel fingerprint, including: Receiving a channel semantic fingerprint sent from at least one distributed node, where the channel semantic fingerprint is semantically compressed channel state information; A pre-trained semantic model is used to perform semantic recovery on the received channel semantic fingerprint to obtain the first channel fingerprint.

10. The device according to claim 8, wherein: The first receiving module is further used to obtain channel state sample data before receiving the channel semantic fingerprint sent from at least one distributed node; The first model to be trained is trained according to the channel state sample data to obtain the semantic model.

11. The device according to claim 10, wherein: The first receiving module trains the first to-be-trained model according to the channel state sample data to obtain the semantic model, including: Extracting features from the channel state sample data to obtain key features; Converting the key features into an image format to obtain a two-dimensional signal fingerprint image set; The first model to be trained is trained according to the two-dimensional signal fingerprint image set until the model converges to obtain the semantic model.

12. The device according to claim 8, wherein: The constructing module constructs a multi-dimensional channel fingerprint matrix according to the first channel fingerprint and the second channel fingerprint, including: The first channel fingerprint and the second channel fingerprint are combined to obtain the multi-dimensional channel fingerprint matrix.

13. The device according to claim 8, wherein: The authentication module authenticates the legitimacy of the transmission channel using a pre-trained identity authentication model according to the multi-dimensional channel fingerprint matrix to obtain a legitimacy authentication result, including: The multi-dimensional channel fingerprint matrix is ​​input into a pre-trained identity authentication model, and the legitimacy of the transmission channel is authenticated through the identity authentication model to obtain a legitimacy authentication result.

14. The device according to any one of claims 8 to 13, wherein: The device also includes a training module, which is used to receive a channel semantic fingerprint sent by at least one distributed node and obtain a distributed node channel semantic fingerprint; Receive the channel fingerprint sent by the sender and obtain the original channel fingerprint; The second model to be trained is trained according to the distributed node channel fingerprint and the original channel fingerprint until the model converges to obtain an identity authentication model.

15. 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 according to any one of claims 1 to 7.

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

17. A computer program product, comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images