Physical layer authentication and channel semantic fingerprint knowledge base construction method and device

Abstract

The invention provides a physical layer authentication method and device, and relates to the technical field of communication, in particular to the technical fields of large models, deep learning, knowledge maps and the like. The specific implementation scheme is as follows: acquiring channel transmission information of a sender in a physical layer; based on the channel transmission information and a pre-constructed channel semantic fingerprint knowledge base, obtaining and deploying a semantic extraction model and an authentication model; obtaining a target semantic fingerprint based on the channel transmission information and the semantic extraction model; and obtaining an authentication result of the sender based on the target semantic fingerprint and the authentication model.

Description

Technical Field

[0001] The present disclosure relates to the computer field, specifically to technical fields such as large models, deep learning, and image processing, and in particular to a physical layer authentication method and device, a channel semantic fingerprint knowledge base construction method and device, an electronic device, and a computer-readable storage medium. Background Art

[0002] With the rapid progress of intelligent wireless communications, semantic communication based on artificial intelligence has gradually become an emerging paradigm in the Internet of Things and communication networks, showing its unique advantages in information transmission and processing. Especially in the 6G era, with the connection of massive devices and the complexity of wireless signals, the demand for physical layer authentication (PLA) has become more prominent. Traditional physical layer authentication methods rely on a large amount of computing resources to extract channel fingerprint features or transmit the channel state information (CSI) of the authenticated user, which can no longer meet the needs in the high-speed and dynamically changing 6G environment. Summary of the invention

[0003] The present disclosure provides a physical layer authentication method and device, a physical layer authentication system, a splicing model training method and device, an electronic device, a computer-readable storage medium, and a computer program product.

[0004] According to the first aspect, a physical layer authentication method is provided, which includes: obtaining channel transmission information of a sender in the physical layer; obtaining and deploying a semantic extraction model and an authentication model based on the channel transmission information and a pre-built channel semantic fingerprint knowledge base; obtaining a target semantic fingerprint based on the channel transmission information and the semantic extraction model; and obtaining an authentication result of the sender based on the target semantic fingerprint and the authentication model.

[0005] According to the second aspect, a method for constructing a channel semantic fingerprint knowledge base is provided, the method comprising: obtaining a channel fingerprint data set; obtaining a channel semantic fingerprint set including at least one channel semantic fingerprint based on the channel fingerprint data set and a semantic extraction model; constructing a fingerprint knowledge graph based on the channel semantic fingerprint set and the channel fingerprint data set; training an authentication model based on the channel semantic fingerprint set and the fingerprint knowledge graph to obtain key parameters of the authentication model; associating the key parameters of the channel semantic fingerprint set, the semantic extraction model, and the authentication model to obtain and store associated data in the channel semantic fingerprint knowledge base.

[0006] According to the third aspect, a physical layer authentication device includes: an information acquisition unit, configured to acquire channel transmission information of a sender in a physical layer; a model acquisition unit, configured to acquire and deploy a semantic extraction model and an authentication model based on the channel transmission information and a pre-built channel semantic fingerprint knowledge base; a fingerprint acquisition unit, configured to acquire a target semantic fingerprint based on the channel transmission information and the semantic extraction model; and a result acquisition unit, configured to acquire an authentication result of the sender based on the target semantic fingerprint and the authentication model.

[0007] According to a fourth aspect, a channel semantic fingerprint knowledge base construction device is provided, and the device includes: a set acquisition unit, configured to acquire a channel fingerprint data set; a data acquisition unit, configured to obtain a channel semantic fingerprint set including at least one channel semantic fingerprint based on the channel fingerprint data set and a semantic extraction model; a construction unit, configured to construct a fingerprint knowledge graph based on the channel semantic fingerprint set and the channel fingerprint data set; a training unit, configured to train an authentication model based on the channel semantic fingerprint set and the fingerprint knowledge graph, and obtain key parameters of the authentication model; an association unit, configured to associate the channel semantic fingerprint set, the semantic extraction model, and the key parameters of the authentication model, and obtain and store associated data in the channel semantic fingerprint knowledge base.

[0008] According to the fifth 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 so that the at least one processor can execute the method described in any implementation of the first aspect or the second aspect.

[0009] According to a sixth aspect, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein 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.

[0010] The physical layer authentication method and device provided by the embodiments of the present disclosure first obtain the channel transmission information of the sender in the physical layer; secondly, based on the channel transmission information and the pre-constructed channel semantic fingerprint knowledge base, obtain and deploy the semantic extraction model and the authentication model; then, based on the channel transmission information and the semantic extraction model, obtain the target semantic fingerprint; finally, based on the target semantic fingerprint and the authentication model, obtain the authentication result of the sender. Thus, the constructed channel semantic fingerprint knowledge base is used to store historical communication data under different devices, users, and network environments, providing a reference for analyzing new channel fingerprints and reducing resource consumption; the constructed channel semantic fingerprint knowledge base is used to record the semantic extraction model and the authentication model, realizing the rapid extraction and authentication of the target semantic fingerprint, and improving the efficiency of physical layer authentication.

[0011] 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

[0012] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0013] Figure 1 is a flow chart of an embodiment of a physical layer authentication method according to the present disclosure;

[0014] Figure 2 It is a structural diagram of the physical layer authentication process disclosed in the present invention;

[0015] Figure 3 is a flow chart of an embodiment of a method for constructing a channel semantic fingerprint knowledge base according to the present disclosure;

[0016] Figure 4 It is a structural schematic diagram of the process of constructing the channel semantic fingerprint knowledge base disclosed in the present invention;

[0017] Figure 5 is a structural diagram of an embodiment of a physical layer authentication device disclosed in the present invention;

[0018] Figure 6 is a structural schematic diagram of an embodiment of a device for constructing a channel semantic fingerprint knowledge base according to the present disclosure;

[0019] Figure 7 It is a block diagram of an electronic device used to implement the physical layer authentication method or the channel semantic fingerprint knowledge base construction method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0020] 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.

[0021] With the advent of 6G, the openness and vulnerability of 6G networks make physical layer authentication more challenging. Attackers may forge identities or tamper with information, threatening data security. Therefore, how to use semantic communication technology to efficiently process CSI information and achieve high-precision physical layer authentication without increasing resource consumption has become a hot topic in current research.

[0022] For example, Solution 1: A threshold-free multi-attribute physical layer authentication method and related equipment, that is, using multi-antenna technology to obtain the attribute value of the physical layer authentication fingerprint of the transmission channel according to the received signal data, and combining the multi-attributes of the physical layer authentication fingerprint to provide multi-dimensional protection for the wireless communication system, effectively resist the risk of certain attributes being invalidated due to being imitated by attackers, and reduce the probability of authentication errors. It can improve the authentication performance when the signal quality is poor. At the same time, Solution 2 is also proposed: A physical layer authentication method and device based on multi-time slot channel characteristics, that is, according to the physical layer characteristic information of the previous time slot and the status information of the current time slot, the physical layer authentication of the sender and the receiver is performed in each time slot, and the physical layer characteristic information of the current time slot is updated. In this way, high security and reliability of data transmission can be achieved.

[0023] However, the scheme 1 simply performs physical layer verification based on the multi-attribute differences of the physical layer obtained from the received signal. It needs to extract multi-attribute features each time, which consumes a lot of computing resources of the device and cannot effectively meet the resource requirements in actual scenarios. The method of scheme 2 using multi-slot physical layer status information for physical layer authentication can effectively identify attackers and distinguish the attacker's identity, but in dynamic scenarios, due to large channel changes, this scheme only relies on the correlation of timing features, and it is easy to misjudge Alice (the sender of the information) as a non-Alice user. In summary, there is a need for a physical layer authentication method with high accuracy, robustness, and high correlation with the communication itself. At the same time, this method can also avoid consuming too many communication and computing resources during the authentication process.

[0024] In view of the above defects, the present invention proposes a physical layer authentication method, which improves the efficiency of physical layer authentication and avoids excessive consumption of communication and computing resources during the authentication process. Figure 1 A process 100 according to an embodiment of a physical layer authentication method of the present disclosure is shown. The physical layer authentication method comprises the following steps:

[0025] Step 101, obtaining the channel transmission information of the sender in the physical layer.

[0026] In this embodiment, the channel is a communication device that connects the sending end and the receiving end, and transmits the signal from the sending end to the receiving end. The channel is divided into: wireless channel and wired channel according to the transmission medium. Among them, the wireless channel uses the propagation of electromagnetic waves in space to transmit signals, and the wired channel uses artificial photoconductive or optical signal media to transmit signals.

[0027] In this embodiment, the physical layer is located at the bottom layer and is responsible for actual data transmission. It provides transparent bit stream transmission services for the data link layer to ensure that data can be transmitted efficiently and reliably on various physical media.

[0028] In this embodiment, the execution subject on which the physical layer authentication method runs is the receiving party (such as Figure 2 The receiver shown in FIG. 1 is a receiver, and the receiver corresponds to the sender. During the communication between the sender and the receiver through the channel, the sender transmits the channel transmission information V (such as Figure 2 As shown in the figure, the information is sent to the execution entity on which the physical layer authentication method runs, and the channel is responsible for transmitting this information. The characteristics of the channel, such as fading and noise, will affect the transmission quality and reliability of the information.

[0029] In this embodiment, in order to realize the physical layer authentication of the transmission channel between the receiver and the sender, it is necessary to obtain the channel transmission information of the sender in the physical layer, wherein the channel transmission information is the information generated by the channel transmission information, and the channel transmission information includes: channel state information (such as Figure 2 The channel transmission information of the sender is the channel transmission information related to the sender. In order to effectively authenticate the physical layer, the channel transmission information can also include: the context information of the sender related to the communication, such as communication mode, historical behavior, etc. The channel transmission information can be used to perform comprehensive physical layer authentication on the sender.

[0030] Optionally, the channel transmission information may also include: user history information (such as Figure 2 As shown), the user history information is the information of the sender at a historical moment, and the user history information includes: historical channel state information.

[0031] Step 102: Based on the channel transmission information and the pre-built channel semantic fingerprint knowledge base, a semantic extraction model and an authentication model are obtained and deployed.

[0032] In this embodiment, the channel semantic fingerprint knowledge base is a pre-built knowledge base of channel fingerprints, which stores channel fingerprints and key parameters of models related to each channel fingerprint, wherein the key parameters are parameters of the deployed model, and the model can be deployed in the terminal through the key parameters; the channel fingerprint is a unique channel feature extracted by measuring and analyzing various aspects of the channel features (such as location, context, and channel status). Each communication link has its unique channel feature, which can be regarded as a "fingerprint". These channel fingerprint information can be used to authenticate the identities of the communicating parties, thereby providing a higher level of security.

[0033] In this embodiment, the semantic extraction model is a pre-trained model, which is used to extract features of channel transmission information, and semantically annotate the extracted features to obtain a channel semantic fingerprint. The semantic extraction model can be one or more, and each semantic extraction model can correspond to a channel semantic fingerprint. The channel semantic fingerprint knowledge base stores the semantic extraction models corresponding to each channel semantic fingerprint.

[0034] In this embodiment, the authentication model is a pre-trained model, which is used to authenticate the channel semantic fingerprint and obtain an authentication result. The authentication result can be the result of various aspects of information after the sender's multiple information (such as identity, location, and environment) is authenticated. For example, the authentication result includes: identity authentication result and geographic location authentication result. The authentication model can be one or more, and each authentication model can correspond to a channel semantic fingerprint. The channel semantic fingerprint knowledge base stores the authentication model corresponding to each channel semantic fingerprint.

[0035] In this embodiment, the above step 102 includes: performing feature extraction on the channel transmission information to obtain extracted features; performing semantic annotation on the extracted features to obtain an identification semantic fingerprint, matching the identification semantic fingerprint with a channel fingerprint in a pre-constructed channel fingerprint knowledge base to obtain a matching channel fingerprint in the channel fingerprint knowledge base; obtaining key parameters of a semantic extraction model and an authentication model related to the matching channel fingerprint in the channel fingerprint knowledge base, and deploying the semantic extraction model and the authentication model on the current terminal based on the key parameters, wherein the key parameters are parameters for constructing the semantic extraction model and the authentication model, and the semantic extraction model and the authentication model can be directly deployed on the terminal through the key parameters.

[0036] Step 103: Obtain a target semantic fingerprint based on the channel transmission information and the semantic extraction model.

[0037] In this embodiment, the target semantic fingerprint is a channel fingerprint that characterizes the channel semantic features of the current sender user and device, wherein the channel semantic features are channel features with semantic information. By authenticating the target semantic fingerprint, it can be ensured that the sender user is legitimate and authorized, which is of great significance for preventing identity masquerade and fake users.

[0038] In this embodiment, the above step 103 includes: inputting the channel transmission information into the semantic extraction model to obtain the target semantic fingerprint.

[0039] Step 104, based on the target semantic fingerprint and the authentication model, obtain the authentication result of the sender.

[0040] In this embodiment, the authentication model authenticates the input target semantic fingerprint to obtain an authentication result. The authentication result of the authentication model is the result obtained after authenticating the target semantic fingerprint. The authentication result of the authentication model can be the identification result of at least one aspect of the sender's multiple information (such as sender's identity information, sender's location, and sender's environment).

[0041] In this embodiment, when the authentication result of the authentication model is the identification result of one aspect of information, the above step 104 includes: sending the target semantic fingerprint to the authentication model to obtain the authentication result of the sender output by the authentication model. When the authentication result of the authentication model is the identification result of multiple aspects of information, the above step 104 includes: sending the target semantic fingerprint to the authentication model to obtain the authentication result output by the authentication model; matching the authentication result output by the authentication model with the pre-built multiple aspects of information comparison table to obtain the authentication result of the sender. Among them, the information comparison table is the authentication results of multiple authentication users in the above multiple aspects.

[0042] The physical layer authentication method provided by the embodiment of the present disclosure first obtains the channel transmission information of the sender in the physical layer; secondly, based on the channel transmission information and the pre-constructed channel semantic fingerprint knowledge base, obtains and deploys the semantic extraction model and the authentication model; then, based on the channel transmission information and the semantic extraction model, obtains the target semantic fingerprint; finally, based on the target semantic fingerprint and the authentication model, obtains the authentication result of the sender. Thus, the constructed channel semantic fingerprint knowledge base is used to store historical communication data under different devices, users, and network environments, providing a reference for analyzing new channel fingerprints and reducing resource consumption; the constructed channel semantic fingerprint knowledge base is used to record the semantic extraction model and the authentication model, realizing the rapid extraction and authentication of the target semantic fingerprint, and improving the efficiency of physical layer authentication.

[0043] Optionally, the target semantic fingerprint may include: a target device fingerprint. The physical layer authentication method disclosed in the present invention further includes: extracting device information from a fingerprint knowledge graph in a channel semantic fingerprint knowledge base based on the target device fingerprint, wherein the device information is information related to the target semantic fingerprint, and the fingerprint knowledge graph is a knowledge graph pre-deployed in the channel semantic fingerprint knowledge base, inputting the device information into the authentication model, and obtaining an authentication result of the target device output by the authentication model, wherein the target device is the sender's device. By authenticating the target device fingerprint, it can be ensured that the devices of the sender and the receiver are legal and credible, which is of great significance for preventing counterfeiting and forgery of devices.

[0044] In some optional implementations of the present disclosure, the above method also includes: based on the authentication result, matching the channel semantic fingerprints of different authentication user information in the channel semantic fingerprint knowledge base with the target semantic fingerprint to obtain a matching result; based on the matching result, updating the authentication model to obtain an updated authentication model; storing the key parameters of the updated authentication model in the channel semantic fingerprint knowledge base.

[0045] In this embodiment, the target semantic fingerprint is a channel semantic fingerprint. When the target semantic fingerprint matches the channel semantic fingerprint of different authenticated user information, it is determined that the sender corresponding to the target semantic fingerprint is a trusted user, and the matching result is a match.

[0046] In this embodiment, updating the authentication model is a process of retraining the authentication model. When the matching result is a match, the target semantic fingerprint and the corresponding identity information are input into the authentication model, so that the model trains the relevance of the current state as a feature for judging it as the identity (dynamic fine-tuning of the authentication model).

[0047] In this embodiment, the authentication model is updated based on the matching result to obtain an updated authentication model, including: in response to a matching result of mismatch, the channel semantic fingerprint corresponding to the matching result is used as a positive sample and the target semantic fingerprint is used as a negative sample to retrain the authentication model to obtain an updated authentication model; in response to a matching result of match, the channel semantic fingerprint and the target semantic fingerprint corresponding to the matching result are both used as positive samples to retrain the authentication model to obtain an updated authentication model; wherein the true value of the positive sample is positive, and the true value of the negative sample is negative, and by setting the true value of the sample, the authentication model can effectively identify the positive sample and the negative sample.

[0048] The physical layer authentication method provided in this embodiment matches the channel semantic fingerprints of different authentication user information in the channel semantic fingerprint knowledge base with the target semantic fingerprint based on the authentication result to obtain a matching result; based on the matching result, the authentication model is updated to obtain an updated authentication model; the key parameters of the updated authentication model are stored in the channel semantic fingerprint knowledge base, thereby matching the authentication result with the channel semantic fingerprint in the channel semantic fingerprint knowledge base, and updating the authentication model based on the matching result, thereby improving the reliability of the authentication model update.

[0049] In some optional implementations of the present disclosure, the above method also includes: updating the semantic extraction model based on the matching results, channel transmission information and target semantic fingerprint to obtain an updated semantic extraction model, and storing the key parameters of the updated semantic extraction model in the channel semantic fingerprint knowledge base.

[0050] In this embodiment, when the matching result is a match, the channel transmission information and the target semantic fingerprint are incorporated into the positive samples of the training semantic extraction model, and the semantic extraction model is retrained to obtain an updated semantic extraction model.

[0051] Optionally, when the matching result is a mismatch, the channel transmission information and the target semantic fingerprint are incorporated into the negative samples of the training semantic extraction model, and the semantic extraction model is retrained to obtain an updated semantic extraction model.

[0052] The physical layer authentication method provided in this embodiment updates the semantic extraction model based on the matching results to obtain an updated semantic extraction model, and stores the key parameters of the updated semantic extraction model in the channel semantic fingerprint knowledge base. Thus, the semantic extraction model is updated through the authentication results, thereby improving the reliability and accuracy of the semantic extraction model update.

[0053] In some optional implementations of the present disclosure, the above-mentioned channel transmission information includes: geographic location information; based on the channel transmission information and a pre-built channel semantic fingerprint knowledge base, obtaining and deploying a semantic extraction model and an authentication model includes: based on the geographic location information, matching key parameters of the semantic extraction model and the authentication model from the channel semantic fingerprint knowledge base; based on the key parameters, deploying the semantic extraction model and the authentication model on the current terminal.

[0054] In this optional implementation, the geographic location information is the sender's perception of the surrounding environment during the signal exchange process, such as surrounding climate information, surrounding buildings, etc. Figure 2 As shown, the geographic location information in the channel transmission information can be used to obtain the key parameters of the semantic extraction model and the authentication model from the channel semantic fingerprint knowledge base, and the semantic extraction model and the authentication model are deployed in the receiver.

[0055] In this optional implementation, the key parameters are the parameters for deploying the semantic extraction model and the authentication model on the terminal, and the semantic extraction model and the authentication model can be directly deployed through the parameters.

[0056] The method for obtaining and deploying a semantic extraction model and an authentication model provided by this optional implementation obtains key parameters of the semantic extraction model and the authentication model from a channel semantic fingerprint knowledge base based on geographic location information; based on the key parameters, the semantic extraction model and the authentication model are deployed on the current terminal, and the key parameters of the semantic extraction model and the authentication model are matched through the fixed characteristic of the geographic location information, thereby improving the reliability and accuracy of the deployment of the semantic extraction model and the authentication model.

[0057] Optionally, the channel semantic fingerprint knowledge base also stores hyperparameters used in the training process of the semantic extraction model and the authentication model, such as learning rate, batch size, training rounds, etc., to ensure the adjustability and reproducibility of the model. The above-mentioned physical layer authentication method also includes: when the semantic extraction model or the authentication model is updated, the hyperparameters of the semantic extraction model or the authentication model are obtained from the channel semantic fingerprint knowledge base, and the semantic extraction model or the authentication model is updated based on the hyperparameters to obtain an updated semantic extraction model or the authentication model.

[0058] Optionally, the above-mentioned physical layer authentication method also includes: when a channel semantic fingerprint of an abnormal user is detected or an attack behavior of the sender is received, timely marking and updating the user information corresponding to the channel semantic fingerprint in the channel semantic fingerprint knowledge base, an abnormal user refers to a user whose channel characteristics suddenly differ greatly from the characteristics stored in the knowledge base, while other attributes do not change too much; the attack behavior refers to the behavior of Eve disguised as Alice discovered during the authentication process.

[0059] In some optional implementations of the present disclosure, the above-mentioned channel transmission information also includes: channel state information, signal environment information; based on the channel transmission information and the semantic extraction model, obtaining the target semantic fingerprint includes: inputting the signal environment information into the semantic extraction model to obtain the environmental semantic label output by the semantic extraction model; based on the geographic location information, obtaining the geographic location label; using the channel state information, the geographic location label and the environmental semantic label as the target semantic fingerprint.

[0060] In this optional implementation, the channel state information is CSI information, which includes: the local channel state information of the sender and the historical channel state information. Signal environment information is the environment information in which the sender sends information at the physical layer. The environment semantic tag can be used to identify different environment types of multiple types of environments. The geographic location tag can be used to identify different geographic locations among multiple geographic locations.

[0061] The method for obtaining the target semantic fingerprint provided by this optional implementation inputs signal environment information into a semantic extraction model to obtain an environmental semantic label output by the semantic extraction model; obtains a geographical location label based on geographical location information; and uses the channel state information, geographical location label, and environmental semantic label as the target semantic fingerprint, thereby improving the comprehensiveness of the target semantic fingerprint information.

[0062] Optionally, the above-mentioned channel transmission information also includes: channel state information; based on the channel transmission information and the semantic extraction model, obtaining the target semantic fingerprint includes: inputting the channel state information into the semantic extraction model to obtain the channel state label output by the semantic extraction model; based on the geographic location information, obtaining the geographic location label; using the channel state label and the geographic location label as the target semantic fingerprint.

[0063] In some optional implementations of the present disclosure, obtaining the authentication result of the sender based on the target semantic fingerprint and the authentication model includes: inputting the target semantic fingerprint into the authentication model to obtain user information output by the authentication model; and obtaining the authentication result of the sender based on the user information.

[0064] In this optional implementation, the user information is the identity information of the sender, such as the sender's ID, name, etc. The authentication model is a model for identifying the user's identity information. The authentication model identifies the target semantic fingerprint of the sender to obtain the sender's user information. Figure 2 As shown, after the receiver obtains the user information output by the authentication model, it can determine the authentication result of the sender by matching it with the user information of the authenticated user in the channel semantic fingerprint knowledge base.

[0065] The method for obtaining the authentication result of the sender provided by this optional implementation inputs the target semantic fingerprint into the authentication model to obtain the user information output by the authentication model; based on the user information, the authentication result of the sender is obtained, and the authentication result is obtained directly using the user information, thereby improving the reliability of physical layer authentication.

[0066] Figure 3 A process 300 of an embodiment of a method for constructing a channel semantic fingerprint knowledge base according to the present disclosure is shown. The method for constructing a channel semantic fingerprint knowledge base includes the following steps:

[0067] Step 301: Acquire a channel fingerprint data set.

[0068] In this embodiment, the channel fingerprint data set may include at least one channel fingerprint data, and the channel fingerprint data includes: local CSI data of the authenticated user and geographical location information of the corresponding signal source (such as Figure 4 As shown), historical CSI data and signal environment information, each channel fingerprint data corresponds to an authenticated user, and each authenticated user has user information.

[0069] Optionally, in order to effectively identify each authenticated user in the channel fingerprint dataset, such as Figure 4 As shown, user behavior features can be added to the channel fingerprint dataset, and each user behavior feature corresponds to an authenticated user. It should be noted that the geographic location information in the channel fingerprint dataset can be directly identified by identification, and the geographic location tag can be obtained and stored in the channel semantic fingerprint knowledge base.

[0070] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of the channel fingerprint data set involved are carried out after authorization and comply with relevant laws and regulations.

[0071] Step 302: obtaining a channel semantic fingerprint set including at least one channel semantic fingerprint based on the channel fingerprint data set and the semantic extraction model.

[0072] In this embodiment, a semantic extraction model is used to perform feature extraction and semantic annotation on each channel fingerprint data in a channel fingerprint data set to obtain a channel semantic fingerprint corresponding to each channel fingerprint data, and a channel semantic fingerprint set is obtained corresponding to the channel fingerprint data set.

[0073] like Figure 4 As shown, the semantic extraction model can be data directly trained by CSI data in the channel fingerprint dataset.

[0074] Step 303: construct a fingerprint knowledge graph based on the channel semantic fingerprint set and the channel fingerprint data set.

[0075] In this embodiment, a relational database (such as MySQL, PostgreSQL) is used to store the channel semantic fingerprint set. Each channel semantic fingerprint in the channel semantic fingerprint set is stored in the form of a table, where each row represents a channel semantic fingerprint.

[0076] In this embodiment, a fingerprint knowledge graph is constructed based on the channel semantic fingerprint, and the channel semantic fingerprint is associated with corresponding equipment, user, environment and other information in the fingerprint knowledge graph.

[0077] Step 304: Based on the channel semantic fingerprint set and the fingerprint knowledge graph, train the authentication model to obtain key parameters of the authentication model.

[0078] In this embodiment, each semantic channel fingerprint in the channel semantic fingerprint set and the identity information corresponding to the semantic channel fingerprint are input into the authentication model, so that the authentication model trains the relevance of the current state as a feature to determine whether it is the identity and gives an authentication result.

[0079] In this embodiment, the above step 304 includes: based on each channel semantic fingerprint in the channel semantic fingerprint set, obtaining authentication information related to each channel semantic fingerprint (including user information of the authenticated user) from the fingerprint knowledge graph, using each channel semantic fingerprint and related authentication information as a sample, training the authentication model, obtaining a trained authentication model, and obtaining key parameters of the authentication model. Among them, the key parameters are parameters for deploying the model, and the model can be deployed in the terminal through the key parameters.

[0080] Optionally, since user behavior characteristics can represent each authenticated user, such as Figure 4 As shown, the authentication model can also be trained through user behavior characteristics.

[0081] Step 305 , associating the channel semantic fingerprint set, the semantic extraction model, and the key parameters of the authentication model, obtaining and storing the associated data in the channel semantic fingerprint knowledge base.

[0082] In this embodiment, after obtaining the semantic extraction model, the key parameters of the semantic extraction model can be determined based on the installation information of the semantic extraction model; the channel semantic fingerprint set, the key parameters of the semantic extraction model, and the key parameters of the authentication model are associated together and stored in the channel semantic fingerprint knowledge base.

[0083] In this embodiment, in order to achieve rapid access to the channel semantic fingerprint knowledge base later, when constructing the channel semantic fingerprint knowledge base, the channel semantic fingerprint knowledge base can be pre-divided into different layers, each layer stores a type of data, and a layer stores the association information of different types of data in each layer. For example, the channel semantic fingerprint knowledge base is divided into: fingerprint layer, semantic layer, authentication layer and public layer, wherein the fingerprint layer is used to store the channel semantic fingerprint set, the semantic layer is used to store the key parameters of the semantic extraction model, the authentication layer is used to store the key parameters of the authentication extraction model, and the public layer is used to store index values ​​related to the associated data, wherein the index value is used to characterize the association relationship between the channel semantic fingerprint set, the key parameters of the semantic extraction model and the key parameters of the authentication extraction model.

[0084] In this embodiment, by constructing a channel semantic fingerprint knowledge base, the model and its parameters, the physical information of the device, and the user's historical characteristics are stored and updated. By adding environmental semantic tags to the data, the relevance to the environment is greatly improved. By using the semantic channel fingerprint knowledge base for physical layer authentication, multi-dimensional fingerprint information authentication is achieved, and the accuracy of authentication is improved. By continuously updating the data and model parameters, the real-time accuracy of the authentication model is guaranteed.

[0085] The channel semantic fingerprint knowledge base construction method provided in this embodiment is as follows: first, a channel fingerprint data set is obtained; second, based on the channel fingerprint data set and the semantic extraction model, a channel semantic fingerprint set including at least one channel semantic fingerprint is obtained; third, based on the channel semantic fingerprint set and the channel fingerprint data set, a fingerprint knowledge graph is constructed; third, based on the channel semantic fingerprint set and the fingerprint knowledge graph, an authentication model is trained to obtain key parameters of the authentication model; finally, the channel semantic fingerprint set, the semantic extraction model, and the key parameters of the authentication model are associated to obtain and store associated data in the channel semantic fingerprint knowledge base, thereby the channel semantic fingerprint, the fingerprint knowledge graph, and the key parameters of the authentication model are associated and stored in the channel semantic fingerprint knowledge base, which enriches the information content of the channel semantic fingerprint knowledge base. When the channel semantic fingerprint knowledge base is used for physical layer authentication, semantic features can be effectively used to extract context information related to communication, thereby improving the accuracy of identity authentication.

[0086] In some optional implementations of the present disclosure, the above-mentioned acquisition of the channel fingerprint data set includes: acquiring a local channel state set of different authenticated user information, and the geographical location of each local channel state in the local channel state set; based on the local channel state set, obtaining a historical channel state set and signal environment information, and using the local channel state information, the historical channel state set, the signal environment information and the geographical location as the channel fingerprint data set.

[0087] In this optional implementation, the local channel state set is a set of channel state information at the current moment, including at least one channel state information, each of which corresponds to an authenticated user. The historical channel state set is a set of channel state information at a historical moment, also including at least one channel state information, each of which corresponds to an authenticated user.

[0088] The method for obtaining a channel fingerprint data set provided by this optional implementation method obtains a local channel state set of different authenticated user information and the geographical location of each local channel state in the local channel state set; based on the local channel state set, a historical channel state set and signal environment information are obtained, and the local channel state information, the historical channel state set, the signal environment information and the geographical location are used as the channel fingerprint data set, which provides a reliable implementation method for obtaining the channel fingerprint data set and improves the comprehensiveness of the channel fingerprint data set information.

[0089] Optionally, the above-mentioned acquisition of the channel fingerprint data set includes: acquiring a local channel state set of different authenticated user information, and the geographical location of each local channel state in the local channel state set; based on the local channel state set, obtaining a historical channel state set, and using the local channel state information, the historical channel state set and the geographical location as the channel fingerprint data set.

[0090] In some optional implementations of the present disclosure, the above-mentioned channel semantic fingerprint set including at least one channel semantic fingerprint based on the channel fingerprint data set and the semantic extraction model includes: inputting the signal environment information in the channel fingerprint data into the semantic extraction model to obtain the environmental semantic label output by the semantic extraction model; obtaining the geographical location label based on the geographical location information in the channel fingerprint data; adding the environmental semantic label and the geographical location label to the channel state in the channel fingerprint data set to obtain the channel semantic fingerprint set including at least one channel semantic fingerprint.

[0091] The method for obtaining a channel semantic fingerprint set provided by this optional implementation inputs signal environment information in the channel fingerprint data into a semantic extraction model to obtain an environmental semantic label output by the semantic extraction model; obtains a geographical location label based on geographical location information in the channel fingerprint data; adds environmental semantic labels and geographical location labels to the channel states in the channel fingerprint data set to obtain a channel semantic fingerprint set including at least one channel semantic fingerprint, thereby improving the comprehensiveness of the channel semantic fingerprint set information.

[0092] In some optional implementations of the present disclosure, the above-mentioned construction of a fingerprint knowledge graph based on a channel semantic fingerprint set and a channel fingerprint data set includes: determining the device, authenticated user information and environment of each channel semantic fingerprint in the channel semantic fingerprint set from the channel fingerprint data set; establishing the relationship between each channel semantic fingerprint, device, authenticated user information and environment, and using corresponding attributes to label the relationship to obtain a fingerprint knowledge graph.

[0093] The method for constructing a fingerprint knowledge graph provided by this optional implementation method determines the device, authenticated user information and environment of each channel semantic fingerprint in a channel semantic fingerprint set from the channel fingerprint data set; establishes the relationship between each channel semantic fingerprint, device, authenticated user information and environment, and uses corresponding attributes to annotate the relationship to obtain a fingerprint knowledge graph, thereby improving the reliability and accuracy of the fingerprint knowledge graph.

[0094] In some optional implementations of the present disclosure, the above-mentioned authentication model is trained based on the channel semantic fingerprint set and the fingerprint knowledge graph to obtain the key parameters of the authentication model, including: extracting the authentication user information related to each information semantic fingerprint in the channel semantic fingerprint set in the fingerprint knowledge graph; inputting each information semantic fingerprint and the corresponding authentication user information into a pre-built authentication model to obtain identity feature information output by the authentication model; based on the identity feature information, calculating the loss value of the authentication model; in response to determining that the authentication model meets the training completion condition based on the loss value, obtaining a trained authentication model.

[0095] In this optional implementation, the loss value of the authentication model can be calculated by using a loss function pre-established for the authentication model; specifically, the above-mentioned calculation of the loss value of the authentication model based on the identity feature information includes: determining a cross-entropy loss function pre-defined for the authentication model, and then calling the cross-entropy loss function and passing in the true label and identity feature information to calculate the loss value of the authentication model.

[0096] In this optional implementation, the training completion condition includes: the loss value of the authentication model is less than a first loss value threshold. The first loss threshold can be determined based on specific training requirements, for example, the first loss threshold is 0.01.

[0097] Optionally, in response to the authentication model not meeting the training completion conditions, the relevant parameters in the authentication model are adjusted so that the loss value of the authentication model converges, and based on the adjusted authentication model, the loss value of the authentication model continues to be calculated, and based on the loss value, it is detected whether the authentication model meets the training completion conditions.

[0098] The method for training an authentication model provided by this optional implementation method extracts authentication user information related to each information semantic fingerprint in a channel semantic fingerprint set in a fingerprint knowledge graph; inputs each information semantic fingerprint and the corresponding authentication user information into a pre-built authentication model to obtain identity feature information output by the authentication model; based on the identity feature information, calculates the loss value of the authentication model; in response to determining that the authentication model meets the training completion conditions based on the loss value, a trained authentication model is obtained, thereby improving the reliability of authentication model training.

[0099] Further references Figure 5 As an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of a physical layer authentication device, which is similar to Figure 1 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0100] like Figure 5 As shown, the physical layer authentication device 500 provided in this embodiment includes: an information acquisition unit 501, a model acquisition unit 502, a fingerprint acquisition unit 503, and a result acquisition unit 504. Among them, the above-mentioned acquisition unit 501 can be configured to acquire the channel transmission information of the sender in the physical layer. The above-mentioned model acquisition unit 502 can be configured to obtain and deploy a semantic extraction model and an authentication model based on the channel transmission information and a pre-built channel semantic fingerprint knowledge base. The above-mentioned fingerprint acquisition unit 503 can be configured to obtain a target semantic fingerprint based on the channel transmission information and the semantic extraction model. The above-mentioned result acquisition unit 504 can be configured to obtain the authentication result of the sender based on the target semantic fingerprint and the authentication model.

[0101] In this embodiment, in the physical layer authentication device 500, the specific processing of the information acquisition unit 501, the model acquisition unit 502, the fingerprint acquisition unit 503, and the result acquisition unit 504 and the technical effects thereof can be referred to respectively. Figure 1 The relevant descriptions of step 101, step 102, step 103, and step 104 in the corresponding embodiment are not repeated here.

[0102] In some optional implementations of the present disclosure, the above-mentioned device also includes: a matching unit (not shown in the figure), and the above-mentioned matching unit is configured to: based on the authentication result, match the channel semantic fingerprints of different authentication user information in the channel semantic fingerprint knowledge base with the target semantic fingerprint to obtain a matching result; based on the matching result, update the authentication model to obtain an updated authentication model; store the key parameters of the updated authentication model in the channel semantic fingerprint knowledge base.

[0103] In some optional implementations of the present disclosure, the above-mentioned device also includes: an updating unit (not shown in the figure), and the above-mentioned updating unit is configured to: update the semantic extraction model based on the matching results, channel transmission information and target semantic fingerprint, obtain an updated semantic extraction model, and store the key parameters of the updated semantic extraction model in the channel semantic fingerprint knowledge base.

[0104] In some optional implementations of the present disclosure, the above-mentioned channel transmission information includes: geographic location information; the above-mentioned model acquisition unit 502 is configured to: based on the geographic location information, obtain key parameters of the semantic extraction model and the authentication model from the channel semantic fingerprint knowledge base by matching; based on the key parameters, deploy the semantic extraction model and the authentication model on the current terminal.

[0105] In some optional implementations of the present disclosure, the above-mentioned channel transmission information also includes: channel state information, signal environment information; the above-mentioned fingerprint acquisition unit 503 is configured to: input the signal environment information into the semantic extraction model to obtain the environmental semantic label output by the semantic extraction model; obtain the geographical location label based on the geographical location information; and use the channel state information, the geographical location label and the environmental semantic label as the target semantic fingerprint.

[0106] In some optional implementations of the present disclosure, the above-mentioned result obtaining unit 504 is configured to: input the target semantic fingerprint into the authentication model to obtain the user information output by the authentication model; and obtain the authentication result of the sender based on the user information.

[0107] The physical layer authentication method provided by the embodiment of the present disclosure is as follows: first, the information acquisition unit 501 acquires the channel transmission information of the sender in the physical layer; secondly, the model acquisition unit 502 obtains and deploys the semantic extraction model and the authentication model based on the channel transmission information and the pre-constructed channel semantic fingerprint knowledge base; then, the fingerprint acquisition unit 503 obtains the target semantic fingerprint based on the channel transmission information and the semantic extraction model; finally, the result acquisition unit 504 obtains the authentication result of the sender based on the target semantic fingerprint and the authentication model. Thus, the constructed channel semantic fingerprint knowledge base is used to store historical communication data under different devices, users, and network environments, providing a reference for analyzing new channel fingerprints and reducing resource consumption; the constructed channel semantic fingerprint knowledge base is used to record the semantic extraction model and the authentication model, realizing the rapid extraction and authentication of the target semantic fingerprint, and improving the efficiency of physical layer authentication.

[0108] Further references Figure 6 As an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of a channel semantic fingerprint knowledge base construction device, which is similar to Figure 3 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0109] like Figure 6 As shown, the channel semantic fingerprint knowledge base construction 600 provided in this embodiment includes: a set acquisition unit 601, a data acquisition unit 602, a construction unit 603, a training unit 604, and an association unit 605. Among them, the above-mentioned set acquisition unit 601 can be configured to acquire a channel fingerprint data set. The above-mentioned data acquisition unit 602 can be configured to obtain a channel semantic fingerprint set including at least one channel semantic fingerprint based on the channel fingerprint data set and the semantic extraction model. The above-mentioned construction unit 603 is configured to construct a fingerprint knowledge graph based on the channel semantic fingerprint set and the channel fingerprint data set. The above-mentioned training unit 604 can be configured to train the authentication model based on the channel semantic fingerprint set and the fingerprint knowledge graph to obtain the key parameters of the authentication model. The above-mentioned association unit 605 can be configured to associate the key parameters of the channel semantic fingerprint set, the semantic extraction model, and the authentication model to obtain and store the associated data in the channel semantic fingerprint knowledge base.

[0110] In this embodiment, the specific processing of the set acquisition unit 601, the data acquisition unit 602, the construction unit 603, the training unit 604, and the association unit 605 and the technical effects thereof can be referred to respectively. Figure 3 The relevant descriptions of step 301, step 302, step 303, step 304, and step 305 in the corresponding embodiments are not repeated here.

[0111] In some optional implementations of the present disclosure, the above-mentioned set acquisition unit 601 is configured to: obtain a local channel state set of different authenticated user information and a geographical location of each local channel state in the local channel state set; based on the local channel state set, obtain a historical channel state set and signal environment information, and use the local channel state information, historical channel state set, signal environment information and geographical location as a channel fingerprint data set.

[0112] In some optional implementations of the present disclosure, the data acquisition unit 602 is configured to: input the signal environment information in the channel fingerprint data into the semantic extraction model to obtain the environmental semantic label output by the semantic extraction model; obtain the geographical location label based on the geographical location information in the channel fingerprint data; add the environmental semantic label and the geographical location label to the channel state in the channel fingerprint data set to obtain a channel semantic fingerprint set including at least one channel semantic fingerprint.

[0113] In some optional implementations of the present disclosure, the above-mentioned construction unit 603 is configured to: determine the device, authenticated user information and environment of each channel semantic fingerprint in the channel semantic fingerprint set from the channel fingerprint data set; establish the relationship between each channel semantic fingerprint, device, authenticated user information and environment, and use corresponding attributes to annotate the relationship to obtain a fingerprint knowledge graph.

[0114] In some optional implementations of the present disclosure, the training unit 604 is configured to: extract authentication user information related to each information semantic fingerprint in the channel semantic fingerprint set in the fingerprint knowledge graph; input each information semantic fingerprint and the corresponding authentication user information into a pre-built authentication model to obtain identity feature information output by the authentication model; based on the identity feature information, calculate the loss value of the authentication model; in response to determining that the authentication model meets the training completion condition based on the loss value, obtain a trained authentication model.

[0115] The channel semantic fingerprint knowledge base construction device provided in this embodiment is as follows: first, the set acquisition unit 601 acquires a channel fingerprint data set; secondly, the data acquisition unit 602 obtains a channel semantic fingerprint set including at least one channel semantic fingerprint based on the channel fingerprint data set and the semantic extraction model; thirdly, the construction unit 603 constructs a fingerprint knowledge graph based on the channel semantic fingerprint set and the channel fingerprint data set; thirdly, the training unit 604 trains an authentication model based on the channel semantic fingerprint set and the fingerprint knowledge graph to obtain key parameters of the authentication model; finally, the association unit 605 associates the channel semantic fingerprint set, the semantic extraction model, and the key parameters of the authentication model to obtain and store the associated data in the channel semantic fingerprint knowledge base, thereby, the channel semantic fingerprint, the fingerprint knowledge graph, and the key parameters of the authentication model are associated and stored in the channel semantic fingerprint knowledge base, enriching the information content of the channel semantic fingerprint knowledge base, and when the channel semantic fingerprint knowledge base is used for physical layer authentication, the semantic features can be effectively used to extract context information related to communication, thereby improving the accuracy of identity authentication.

[0116] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information (such as authenticated user information or user information of authenticated users) involved shall comply with the relevant laws and regulations and shall not violate public order and good morals.

[0117] 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.

[0118] Figure 7 A schematic block diagram of an example electronic device 700 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.

[0119] like Figure 7As shown, the device 700 includes a computing unit 701, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 can also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0120] A number of components in the device 700 are connected to the I / O interface 705, including: an input unit 706, such as a keyboard, a mouse, etc.; an output unit 707, such as various types of displays, speakers, etc.; a storage unit 708, such as a disk, an optical disk, etc.; and a communication unit 709, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 709 allows the device 700 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0121] The computing unit 701 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 701 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 701 performs the various methods and processes described above, such as a physical layer authentication method or a channel semantic fingerprint knowledge base construction method. For example, in some embodiments, the physical layer authentication method or the channel semantic fingerprint knowledge base construction method may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 708. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 700 via the ROM 702 and / or the communication unit 709. When the computer program is loaded into the RAM 703 and executed by the computing unit 701, one or more steps of the physical layer authentication method or the channel semantic fingerprint knowledge base construction method described above may be performed. Alternatively, in other embodiments, the computing unit 701 may be configured to execute the physical layer authentication method or the channel semantic fingerprint knowledge base construction method in any other appropriate manner (for example, by means of firmware).

[0122] Various implementations of the systems and techniques described above 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), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including 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.

[0123] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable physical layer authentication device or a channel semantic fingerprint knowledge base construction 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 may be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0124] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may 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 may 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.

[0125] 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).

[0126] 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 information 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 communications network). Examples of communications networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0127] 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.

[0128] 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.

[0129] 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 physical layer authentication method, the method comprising: Obtain the channel transmission information of the sender in the physical layer; Based on the channel transmission information and a pre-built channel semantic fingerprint knowledge base, a semantic extraction model and an authentication model are obtained and deployed; Obtaining a target semantic fingerprint based on the channel transmission information and the semantic extraction model; Based on the target semantic fingerprint and the authentication model, an authentication result of the sender is obtained.

2. The method according to claim 1, further comprising: Based on the authentication result, matching the channel semantic fingerprints of different authenticated users in the channel semantic fingerprint knowledge base with the target semantic fingerprint to obtain a matching result; Based on the matching result, the authentication model is updated to obtain an updated authentication model; The key parameters of the updated authentication model are stored in the channel semantic fingerprint knowledge base.

3. The method according to claim 1 or 2, further comprising: Based on the matching result, the channel transmission information and the target semantic fingerprint, the semantic extraction model is updated to obtain an updated semantic extraction model, and key parameters of the updated semantic extraction model are stored in a channel semantic fingerprint knowledge base.

4. The method according to claim 1 or 2, wherein: The channel transmission information includes: geographic location information; the obtaining and deploying of a semantic extraction model and an authentication model based on the channel transmission information and a pre-built channel semantic fingerprint knowledge base includes: Based on the geographic location information, key parameters of a semantic extraction model and an authentication model are obtained by matching from the channel semantic fingerprint knowledge base; Based on the key parameters, the semantic extraction model and the authentication model are deployed on the current terminal.

5. The method according to claim 4, wherein: The channel transmission information further includes: channel state information and signal environment information; and obtaining a target semantic fingerprint based on the channel transmission information and the semantic extraction model includes: Inputting the signal environment information into the semantic extraction model to obtain an environment semantic label output by the semantic extraction model; Based on the geographic location information, obtain a geographic location tag; The channel state information, the geographic location tag, and the environmental semantic tag are used as target semantic fingerprints.

6. The method according to claim 1 or 2, wherein: The obtaining the authentication result of the sender based on the target semantic fingerprint and the authentication model includes: Inputting the target semantic fingerprint into the authentication model to obtain user information output by the authentication model; Based on the user information, an authentication result of the sender is obtained.

7. A method for constructing a channel semantic fingerprint knowledge base, the method comprising: Obtain a channel fingerprint dataset; Based on the channel fingerprint data set and the semantic extraction model, obtaining a channel semantic fingerprint set including at least one channel semantic fingerprint; Based on the channel semantic fingerprint set and the channel fingerprint data set, construct a fingerprint knowledge graph; Based on the channel semantic fingerprint set and the fingerprint knowledge graph, training an authentication model to obtain key parameters of the authentication model; The channel semantic fingerprint set, the semantic extraction model, and key parameters of the authentication model are associated to obtain and store associated data in a channel semantic fingerprint knowledge base.

8. The method according to claim 7, wherein: The acquiring of a channel fingerprint data set comprises: Acquire a local channel state set of different authenticated users and a geographical location of each local channel state in the local channel state set; Based on the local channel state set, a historical channel state set and signal environment information are obtained, and the local channel state information, the historical channel state set, the signal environment information and the geographical location are used as a channel fingerprint data set.

9. The method according to claim 8, wherein: The step of obtaining a channel semantic fingerprint set including at least one channel semantic fingerprint based on the channel fingerprint data set and the semantic extraction model comprises: Inputting the signal environment information in the channel fingerprint data into the semantic extraction model to obtain the environment semantic label output by the semantic extraction model; Obtaining a geographic location tag based on the geographic location information in the channel fingerprint data; An environmental semantic tag and a geographical location tag are added to the channel state in the channel fingerprint data set to obtain a channel semantic fingerprint set including at least one channel semantic fingerprint.

10. The method according to claim 7, wherein: The constructing of a fingerprint knowledge graph based on the channel semantic fingerprint set and the channel fingerprint data set includes: Determine, from the channel fingerprint data set, the device, the authenticated user information, and the environment of each channel semantic fingerprint in the channel semantic fingerprint set; Establish the relationship between each channel semantic fingerprint, device, authenticated user information and environment, and use the corresponding attributes to annotate the relationship to obtain the fingerprint knowledge graph.

11. The method according to claim 7, wherein: The authentication model is trained based on the channel semantic fingerprint set and the fingerprint knowledge graph to obtain key parameters of the authentication model, including: Extracting authentication user information related to each information semantic fingerprint in the channel semantic fingerprint set from the fingerprint knowledge graph; Inputting each information semantic fingerprint and corresponding authentication user information into a pre-built authentication model to obtain identity feature information output by the authentication model; Based on the identity feature information, calculating the loss value of the authentication model; In response to determining that the certification model meets the training completion condition based on the loss value, a training completed certification model is obtained.

12. A physical layer authentication device, the device comprising: An information acquisition unit, configured to acquire channel transmission information of a sender in a physical layer; A model obtaining unit is configured to obtain and deploy a semantic extraction model and an authentication model based on the channel transmission information and a pre-built channel semantic fingerprint knowledge base; A fingerprint obtaining unit, configured to obtain a target semantic fingerprint based on the channel transmission information and the semantic extraction model; The result obtaining unit is configured to obtain the authentication result of the sender based on the target semantic fingerprint and the authentication model.

13. A channel semantic fingerprint knowledge base construction device, the device comprising: A set acquisition unit is configured to acquire a channel fingerprint data set; A data acquisition unit, configured to obtain a channel semantic fingerprint set including at least one channel semantic fingerprint based on the channel fingerprint data set and the semantic extraction model; A construction unit, configured to construct a fingerprint knowledge graph based on the channel semantic fingerprint set and the channel fingerprint data set; A training unit is configured to train an authentication model based on the channel semantic fingerprint set and the fingerprint knowledge graph to obtain key parameters of the authentication model; The associating unit is configured to associate the channel semantic fingerprint set, the semantic extraction model, and key parameters of the authentication model to obtain and store associated data in a channel semantic fingerprint knowledge base.

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 11.

15. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to enable the computer to execute the method according to any one of claims 1 to 11.

Images