Multi-device radio frequency fingerprint synchronous identification model training and identification method

Abstract

The invention provides a multi-device radio frequency fingerprint synchronous identification model training and identification method, the multi-device radio frequency fingerprint synchronous identification model comprises a first sub-model and a second sub-model, and the training method comprises the following steps: obtaining training data; performing signal preprocessing and multipath enhancement on the plurality of training signals to obtain a plurality of sample signals; inputting the plurality of sample signals into a first initial sub-model to obtain a plurality of first tensors; inputting each first tensor into a second initial sub-model to obtain a second tensor output by the second initial sub-model; and obtaining a loss function based on the second tensor and the link identifier, and performing parameter adjustment on the first initial sub-model and the second initial sub-model based on the loss function to obtain the first sub-model and the second sub-model. According to the scheme of the invention, accurate multi-device radio frequency fingerprint synchronous identification can be carried out in a multi-antenna multi-user scene.

Description

Technical Field

[0001] The present application relates to the field of radio frequency fingerprint recognition technology, and in particular to a multi-device radio frequency fingerprint synchronous recognition model training and recognition method. Background Art

[0002] Radio frequency fingerprinting is an identification technology based on the hardware characteristics of radio frequency signals from wireless devices. In highly dynamic communication scenarios (e.g., WiFi communications), the channel state between communication devices typically changes frequently. In such scenarios, radio frequency fingerprinting systems based on wireless channels need to constantly update the channel state fingerprint between the authenticated device and the authenticating device, resulting in significant communication and computational overhead. Radio frequency fingerprinting technologies based on hardware characteristics do not have the aforementioned drawbacks, but most related technologies only target single-antenna, single-link scenarios. Summary of the Invention

[0003] The present application aims to solve one of the technical problems in the related art at least to a certain extent.

[0004] In the first aspect, the present application proposes a training method for a multi-device radio frequency fingerprint synchronous recognition model, the method comprising: obtaining training data; wherein the training data comprises multiple training signals and link identifiers of links corresponding to the multiple training signals; performing signal preprocessing and multipath enhancement on the multiple training signals to obtain multiple sample signals; inputting the multiple sample signals into a first initial sub-model to obtain multiple first tensors; wherein the first initial sub-model is used to process the multiple sample signals to obtain multiple tensors, and perform tensor slicing on the multiple tensors according to the number of links corresponding to the multiple signals to obtain the multiple first tensors; inputting each of the first tensors into a second initial sub-model respectively to obtain a second tensor output by the second initial sub-model; obtaining a loss function based on the second tensor and the link identifier, and adjusting the parameters of the first initial sub-model and the second initial sub-model based on the loss function to obtain the first sub-model and the second sub-model.

[0005] In one implementation, performing signal preprocessing and multipath enhancement on the multiple training signals to obtain multiple sample signals includes: performing time domain to frequency domain transformation and normalization processing on the multiple training signals to obtain corresponding multiple first signals; and performing multipath channel enhancement and noise enhancement on the multiple first signals to obtain the multiple sample signals.

[0006] In an optional implementation, multipath channel enhancement is performed on the multiple first signals based on the following formula:

[0007]

[0008] Among them, sjl is the first signal at the jth link and the lth symbol before multipath channel enhancement, s ′ kl is the signal at the kth receiving link and the lth symbol after multipath channel enhancement, represents element-wise product, is the frequency domain response of a randomly generated multipath fading channel, N car is the number of subcarriers or frequencies.

[0009] In one implementation, the loss function is calculated as follows:

[0010]

[0011] in, is the loss function, N tx is the number of links, r j is the jth second tensor, Represents r j c j The elements at position c j For r j The corresponding link identifier.

[0012] In one implementation, the first initial sub-model and the second initial sub-model both include a residual module, the residual module includes at least one convolutional layer, and each of the convolutional layers includes a batch normalization layer and a Sigmoid activation layer.

[0013] In the second aspect, the present application proposes a method for synchronous identification of multi-device radio frequency fingerprints, including: obtaining a signal to be identified; wherein the signal to be identified includes multiple sub-signals; performing signal preprocessing and multipath enhancement on the signal to be identified to obtain an input signal; inputting the input signal into a multi-device radio frequency fingerprint synchronous identification model to obtain a link identifier corresponding to each of the sub-signals; wherein the multi-device radio frequency fingerprint synchronous identification model is trained based on the method described in the first aspect; based on the mapping relationship between the target link identifier and the pre-acquired link identifier and the device identifier, performing synchronous identification of multi-device radio frequency fingerprints.

[0014] In a third aspect, the present application proposes a training device for a multi-device radio frequency fingerprint synchronous identification model, wherein the multi-device radio frequency fingerprint synchronous identification model includes a first sub-model and a second sub-model, and the device includes: an acquisition module for acquiring training data; wherein the training data includes multiple training signals and link identifiers of links corresponding to the multiple training signals; a first processing module for performing signal preprocessing and multipath enhancement on the multiple training signals to obtain multiple sample signals; a second processing module for inputting the multiple sample signals into a first initial sub-model to obtain multiple first tensors; wherein the first initial sub-model is used to process the multiple sample signals to obtain multiple tensors, and to perform tensor slicing on the multiple tensors according to the number of links corresponding to the multiple signals to obtain the multiple first tensors; a third processing module for inputting each first tensor into a second initial sub-model respectively to obtain a second tensor output by the second initial sub-model; a fourth processing module for obtaining a loss function based on the second tensor and the link identifier, and adjusting parameters of the first initial sub-model and the second initial sub-model based on the loss function to obtain the first sub-model and the second sub-model.

[0015] In one implementation, the first processing module can be used to: perform time domain to frequency domain transformation and normalization processing on the multiple training signals to obtain corresponding multiple first signals; and perform multipath channel enhancement and noise enhancement on the multiple first signals to obtain the multiple sample signals.

[0016] In an optional implementation, the first processing module may perform multipath channel enhancement on the multiple first signals based on the following formula:

[0017]

[0018] Among them, s jl is the first signal at the jth link and the lth symbol before multipath channel enhancement, s ′ kl is the signal at the kth receiving link and the lth symbol after multipath channel enhancement, ° represents the element-by-element product, is the frequency domain response of a randomly generated multipath fading channel, N car is the number of subcarriers or frequencies.

[0019] In one implementation, the loss function is calculated as follows:

[0020]

[0021] in, is the loss function, N tx is the number of links, r jis the jth second tensor, Represents r j c j The elements at position c j For r j The corresponding link identifier.

[0022] In one implementation, the first initial sub-model and the second initial sub-model both include a residual module, the residual module includes at least one convolutional layer, and each of the convolutional layers includes a batch normalization layer and a Sigmoid activation layer.

[0023] In the fourth aspect, the present application proposes a multi-device radio frequency fingerprint synchronous identification device, including: an acquisition module for acquiring a signal to be identified; wherein the signal to be identified includes multiple sub-signals; a first processing module for performing signal preprocessing and multipath enhancement on the signal to be identified to obtain an input signal; a second processing module for inputting the input signal into a multi-device radio frequency fingerprint synchronous identification model to obtain a link identifier corresponding to each of the sub-signals; wherein the multi-device radio frequency fingerprint synchronous identification model is trained based on the method described in the first aspect; an identification module for performing multi-device radio frequency fingerprint synchronous identification based on the target link identifier and the mapping relationship between the pre-acquired link identifier and the device identifier.

[0024] In a fifth aspect, the present application proposes an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the training method of the multi-device radio frequency fingerprint synchronous recognition model as described in the first aspect.

[0025] In the sixth aspect, the present application proposes an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the multi-device radio frequency fingerprint synchronous identification method as described in the first aspect.

[0026] In a seventh aspect, the present application proposes a computer-readable storage medium for storing instructions, which, when executed, enables the method described in the first aspect to be implemented.

[0027] In an eighth aspect, the present application proposes a computer-readable storage medium for storing instructions, which, when executed, enables the method described in the second aspect to be implemented.

[0028] In a ninth aspect, the present application proposes a computer program product, comprising a computer program, which, when executed by a processor, implements the steps of the training method for the multi-device radio frequency fingerprint synchronous recognition model as described in the first aspect.

[0029] In a tenth aspect, the present application proposes a computer program product, comprising a computer program, which, when executed by a processor, implements the steps of the multi-device radio frequency fingerprint synchronous identification method as described in the first aspect.

[0030] The training method, apparatus, device, and storage medium for a multi-device RF fingerprint synchronous identification model provided in this application can perform signal preprocessing and multipath enhancement on multiple training signals to obtain multiple sample signals, and input the multiple sample signals into a first initial sub-model to obtain multiple first tensors. Each of the first tensors is input into a second initial sub-model to obtain a second tensor output by the second initial sub-model, thereby obtaining a loss function based on the second tensor and the link identifier, and adjusting the parameters of the first initial sub-model and the second initial sub-model based on the loss function to obtain a multi-device RF fingerprint synchronous identification model, so as to perform multi-device RF fingerprint synchronous identification based on the multi-device RF fingerprint synchronous identification model. It can achieve independent extraction and accurate identification of RF fingerprints.

[0031] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0033] Figure 1 This is a flowchart of a method for training a multi-device radio frequency fingerprint synchronous recognition model provided by an embodiment of the present application;

[0034] Figure 2 is a structural diagram of a first initial sub-model provided in an embodiment of the present application;

[0035] Figure 3 is a structural diagram of a second initial sub-model provided in an embodiment of the present application;

[0036] Figure 4 This is a flowchart of another method for training a multi-device radio frequency fingerprint synchronous recognition model provided by an embodiment of the present application;

[0037] Figure 5 This is a flowchart of a multi-device radio frequency fingerprint synchronous identification method provided by an embodiment of the present application;

[0038] Figure 6 This is a structural diagram of a multi-device radio frequency fingerprint synchronous recognition model provided by an embodiment of the present application;

[0039] Figure 7 This is a structural diagram of a training device for a multi-device radio frequency fingerprint synchronous recognition model provided by an embodiment of the present application;

[0040] Figure 8 This is a structural diagram of a multi-device radio frequency fingerprint synchronous identification device proposed in this application;

[0041] Figure 9 It is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0042] The following describes in detail embodiments of the present application, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0043] The following describes a multi-device radio frequency fingerprint synchronous recognition model training and recognition method according to an embodiment of the present application with reference to the accompanying drawings.

[0044] It should be noted that the multi-device radio frequency fingerprint synchronous recognition model training and recognition method provided in the embodiment of the present application can be applied to MIMO (Multiple-Input Multiple-Output) scenarios.

[0045] Figure 1 This is a flow chart of a method for training a multi-device radio frequency fingerprint synchronous recognition model provided by an embodiment of the present application. The multi-device radio frequency fingerprint synchronous recognition model may include a first sub-model and a second sub-model. Figure 1 As shown, the method may include but is not limited to the following steps:

[0046] Step S101: Acquire training data.

[0047] The training data includes multiple training signals and link identifiers of links corresponding to the multiple training signals.

[0048] Exemplarily, the link identifier may be a link ID (Identity document).

[0049] It should be noted that in the embodiments of the present application, there may be multiple training data. Based on each training data, the training method of the multi-device radio frequency fingerprint synchronous recognition model of any embodiment of the present application can be used to perform multiple training to obtain a multi-device radio frequency fingerprint synchronous recognition model.

[0050] For example, each training data may include multiple training signals and links

[0051] It can be understood that the number of training signals in the training data is greater than or equal to the number of links, each training signal corresponds to a link, and the links corresponding to different training signals may be the same or different.

[0052] Step S102: performing signal preprocessing and multipath enhancement on a plurality of training signals to obtain a plurality of sample signals.

[0053] Exemplarily, the training signal is sequentially subjected to time-frequency domain transformation, normalization processing, channel enhancement, and noise enhancement to obtain a corresponding plurality of sample signals.

[0054] Exemplarily, the time-frequency domain transformation may be any one of the following: Fourier transform, Laplace transform, Z transform, and wavelet transform.

[0055] Exemplarily, the above-mentioned normalization processing may be any one of the following: power normalization and constellation mapping normalization.

[0056] Step S103: Input multiple sample signals into the first initial sub-model to obtain multiple first tensors.

[0057] The first initial sub-model is used to process multiple sample signals to obtain multiple tensors, and to perform tensor slicing on the multiple tensors according to the number of links corresponding to the multiple signals to obtain multiple first tensors.

[0058] In one implementation, the first initial sub-model includes a residual module, the residual module includes at least one convolutional layer, and each convolutional layer includes a batch normalization layer and a Sigmoid activation layer.

[0059] For example, see Figure 2 , Figure 2 This is a schematic diagram of the structure of a first initial sub-model provided in an embodiment of the present application. Figure 2 As shown in the figure, the first initial sub-model includes a residual module with a customizable number of stacking times, and all convolutional layers in the first initial sub-model are accompanied by a batch normalization layer and a Sigmoid activation layer.

[0060] For example, one or more sample signals are For example, multiple sample signals are input into the first initial sub-model, and the first initial sub-model first processes the multiple sample signals to obtain multiple tensors Then the first model slices the above multiple tensors according to the number of links corresponding to the multiple sample signals, and outputs tensors with the same number of links. For example, tensors are discarded in the output order. Output tensors with the same number of links

[0061] Step S104: input each first tensor into the second initial sub-model respectively to obtain a second tensor output by the second initial sub-model.

[0062] Among them, each second tensor corresponds to a link ID.

[0063] In one implementation, the second initial sub-model and the second initial sub-model both include a residual module, the residual module includes at least one convolutional layer, and each of the convolutional layers includes a batch normalization layer and a Sigmoid activation layer.

[0064] For example, see Figure 3 , Figure 3 This is a schematic diagram of the structure of a second initial sub-model provided in an embodiment of the present application. Figure 3 As shown in Figure 1, the second initial sub-model includes residual modules that can be stacked, and all convolutional layers in the second initial sub-model are accompanied by a batch normalization layer and a Sigmoid activation layer.

[0065] Exemplarily, each first tensor output by the first initial sub-model based on a sample data is input into the second initial sub-model to obtain a second tensor output by the second initial sub-model.

[0066] Step S105: Obtain a loss function based on the second tensor and the link identifier, and adjust parameters of the first initial sub-model and the second initial sub-model based on the loss function to obtain the first sub-model and the second sub-model.

[0067] In one implementation, the second tensor output by the second initial sub-model is For example, the calculation formula of the loss function can be expressed as follows:

[0068]

[0069] in, is the loss function, N tx is the number of links, r j is the j-th second tensor, Represents r j c j The elements at position c j For rj Corresponding link identifier.

[0070] Exemplarily, the above formula is used to calculate the loss function, and the Adam optimizer is used to perform gradient backpropagation on the first initial sub-model and the second initial sub-model to update the model parameters of the first initial sub-model and the second initial sub-model, and repeated training is performed using different sample data until the loss function value is less than a preset threshold, thereby obtaining the first sub-model and the second sub-model of the multi-device radio frequency fingerprint synchronous recognition model.

[0071] By implementing the embodiments of the present application, multiple training signals can be subjected to signal preprocessing and multipath enhancement to obtain multiple sample signals, and the multiple sample signals can be input into the first initial sub-model to obtain multiple first tensors. Each first tensor is input into the second initial sub-model to obtain a second tensor output by the second initial sub-model. A loss function is then obtained based on the second tensor and the link identifier, and the parameters of the first and second initial sub-models are adjusted based on the loss function to obtain a multi-device RF fingerprint synchronization recognition model. A multi-device RF fingerprint synchronization recognition model can be obtained that can accurately separate and identify RF fingerprints of multiplexed signals in a multi-antenna reception scenario.

[0072] In some embodiments, the training signal in the training data may be subjected to signal preprocessing and multipath enhancement to obtain multiple sample signals. As an example, see Figure 4 , Figure 4 This is a flow chart of another method for training a multi-device radio frequency fingerprint synchronous recognition model provided by an embodiment of the present application. Figure 4 As shown, the method may include but is not limited to the following steps:

[0073] Step S401: Acquire training data.

[0074] In the embodiment of the present application, step S401 can be implemented in any of the ways in the embodiments of the present application. The embodiments of the present application do not limit this and will not be described in detail.

[0075] Step S402: performing time domain to frequency domain transformation and normalization processing on the multiple training signals to obtain corresponding multiple first signals.

[0076] Exemplarily, the training signal is subjected to Fourier transform and then power normalized so that the power of the result is 1, thereby obtaining the corresponding first signal.

[0077] For example, the training signal is For example, the training signal after time domain to frequency domain transformation and power normalization can be expressed as in, is the received signal of the i-th receiving link, which is an Nsym ×N car The complex tensor N sym is the number of symbols in the training field, N car is the number of subcarriers or frequencies.

[0078] Step S403: performing multipath channel enhancement and noise enhancement on the multiple first signals to obtain multiple sample signals.

[0079] In an optional implementation, the first signal is Taking as an example, the above multipath channel enhancement can be expressed as follows:

[0080]

[0081] Among them, s jl To enhance the first signal at the first jth receiving link and the first symbol, s ′ kl For the enhanced signal at the kth receiving link and the lth symbol, represents element-wise product, is the frequency domain response of a randomly generated multipath fading channel, N car is the number of subcarriers or frequencies.

[0082] In some embodiments, the multipath fading channel may be a frequency-flat fading channel or a frequency-selective fading channel.

[0083] Step S404: Input multiple sample signals into the first initial sub-model to obtain a first tensor group.

[0084] In the embodiment of the present application, step S404 can be implemented by any of the methods in the embodiments of the present application. The embodiments of the present application do not limit this and will not be described in detail.

[0085] Step S405: Input each first tensor into the second initial sub-model respectively to obtain a second tensor output by the second initial sub-model.

[0086] In the embodiment of the present application, step S405 can be implemented by any of the methods in the embodiments of the present application. The embodiments of the present application do not limit this and will not be described in detail.

[0087] Step S406: Obtain a loss function based on the second tensor and the link identifier, and adjust parameters of the first initial sub-model and the second initial sub-model based on the loss function to obtain the first sub-model and the second sub-model.

[0088] In the embodiment of the present application, step S406 can be implemented by any of the methods in the embodiments of the present application. The embodiments of the present application do not limit this and will not be described in detail.

[0089] By implementing the embodiments of the present application, the signals in the training data can be preprocessed and multipath enhanced to obtain multiple sample signals, so as to train a multi-device radio frequency fingerprint synchronous recognition model based on the multiple sample signals.

[0090] See Figure 5 , Figure 5 This is a flow chart of a multi-device radio frequency fingerprint synchronous identification method provided by an embodiment of the present application. Figure 5 As shown, the method may include but is not limited to the following steps:

[0091] Step S501: Acquire a signal to be identified.

[0092] The signal to be identified includes multiple sub-signals.

[0093] Exemplarily, the received MIMO signal is used as the signal to be identified.

[0094] Step S502: performing signal preprocessing and multipath enhancement on the signal to be identified to obtain an input signal.

[0095] Exemplarily, the same method as step S402 and step S403 may be used to perform signal preprocessing and multipath enhancement on the signal to be identified to obtain an input signal.

[0096] Step S503: Input the input signal into the multi-device radio frequency fingerprint synchronous recognition model to obtain the target link identifier corresponding to each sub-signal.

[0097] In an embodiment of the present application, the multi-device radio frequency fingerprint synchronous identification model includes a first sub-model and a second sub-model, and the multi-device radio frequency fingerprint synchronous identification model is trained based on the training method of the multi-device radio frequency fingerprint synchronous identification model provided in any embodiment of the present application.

[0098] Exemplarily, the input signal is input into the first sub-model to separate and extract features of the input signal, obtain feature information of the signal corresponding to each transmission link, and input the feature information into the second sub-model so that the second sub-model processes the feature based on the feature to obtain a feature tensor, and performs maximum index calculation based on the feature tensor to output the corresponding target link identifier.

[0099] Step S504: Based on the target link identifier and the pre-acquired mapping relationship between the link identifier and the device identifier, perform multi-device radio frequency fingerprint synchronous identification.

[0100] In some embodiments, the above-mentioned device identification is the device identification of the device sending the training signal in the embodiment of the training method of the multi-device radio frequency fingerprint synchronous recognition model of the present application.

[0101] Exemplarily, the device identification may be a device ID.

[0102] Exemplarily, based on a pre-acquired mapping relationship between link identifiers and transmitting device identifiers, a target device corresponding to each link identifier is determined to perform simultaneous identification of radio frequency fingerprints of multiple devices.

[0103] By implementing the embodiments of the present application, accurate multi-device RF fingerprint synchronous identification can be performed in a multi-user multi-antenna scenario based on the trained multi-device RF fingerprint synchronous identification model.

[0104] For example, please participate Figure 6 , Figure 6 This is a structural diagram of a multi-device radio frequency fingerprint synchronization recognition model provided by an embodiment of the present application. Figure 6 As shown, the multi-device RF fingerprint synchronous identification model includes sub-network model 1 and sub-network model 2. Sub-network model 1 can separate the input signals of multiple transmission links and output the characteristics of multiple transmission links. Sub-network model 2 can perform multi-device RF fingerprint synchronous identification based on the characteristics of each transmission link to obtain the corresponding link ID. Based on the mapping relationship between the link ID and the device ID of the signal transmitting device, the device corresponding to the link is determined, thereby realizing multi-device RF fingerprint synchronous identification.

[0105] See Figure 7 , Figure 7 This is a structural diagram of a training device for a multi-device radio frequency fingerprint synchronous recognition model provided by an embodiment of the present application. The multi-device radio frequency fingerprint synchronous recognition model includes a first sub-model and a second sub-model. Figure 7As shown, the device 700 includes: an acquisition module 701, which is used to acquire training data; wherein the training data includes multiple training signals and link identifiers of links corresponding to the multiple training signals; a first processing module 702, which is used to perform signal preprocessing and multipath enhancement on the multiple training signals to obtain multiple sample signals; a second processing module 703, which is used to input the multiple sample signals into the first initial sub-model to obtain multiple first tensors; wherein the first initial sub-model is used to process the multiple sample signals to obtain multiple tensors, and to perform tensor slicing on the multiple tensors according to the number of links corresponding to the multiple signals to obtain multiple first tensors; a third processing module 704, which is used to input each first tensor into the second initial sub-model respectively to obtain a second tensor output by the second initial sub-model; a fourth processing module 705, which is used to obtain a loss function based on the second tensor and the link identifier, and to adjust the parameters of the first initial sub-model and the second initial sub-model based on the loss function to obtain the first sub-model and the second sub-model.

[0106] In one implementation, the first processing module 702 may be configured to: perform time-domain to frequency-domain transformation and normalization processing on multiple training signals to obtain corresponding multiple first signals; and perform multipath channel enhancement and noise enhancement on the multiple first signals to obtain multiple sample signals.

[0107] In an optional implementation, the first processing module 702 may perform multipath channel enhancement on the multiple first signals based on the following formula:

[0108]

[0109] Among them, s jl The signal at the jth link and the lth symbol before multipath channel enhancement, s ′ kl is the signal at the kth receiving link and the lth symbol after multipath channel enhancement, represents element-wise product, is the frequency domain response of a randomly generated multipath fading channel, N car is the number of subcarriers or frequencies.

[0110] In one implementation, the loss function is calculated as:

[0111]

[0112] in, is the loss function, N tx is the number of links, r j is the j-th second tensor, Represents r j c j The elements at position c jFor r j Corresponding link identifier.

[0113] In one implementation, the first initial sub-model and the second initial sub-model both include a residual module, the residual module includes at least one convolutional layer, and each convolutional layer includes a batch normalization layer and a sigmoid activation layer.

[0114] Through the apparatus of the embodiment of the present application, multiple training signals can be preprocessed and multipath enhanced to obtain multiple sample signals. The multiple sample signals are then input into a first initial sub-model to obtain multiple first tensors. Each first tensor is then input into a second initial sub-model to obtain a second tensor output by the second initial sub-model. The second tensor and the link identifier are then used to obtain a loss function. The parameters of the first and second initial sub-models are then adjusted based on the loss function to obtain a multi-device RF fingerprint simultaneous identification model. This results in a neural network model that can accurately separate multiplexed signals and identify RF fingerprints in multi-antenna reception scenarios.

[0115] It should be noted that the above explanation of the embodiment of the training method of the multi-device radio frequency fingerprint synchronous recognition model is also applicable to the training device of the multi-device radio frequency fingerprint synchronous recognition model of this embodiment, and will not be repeated here.

[0116] See Figure 8 , Figure 8 This is a schematic diagram of the structure of a multi-device radio frequency fingerprint synchronous identification device proposed in this application. Figure 8 As shown, the device 800 includes: an acquisition module 801, used to obtain a signal to be identified; wherein the signal to be identified includes multiple sub-signals; a first processing module 802, used to perform signal preprocessing and multipath enhancement on the signal to be identified to obtain an input signal; a second processing module 803, used to input the input signal into a multi-device radio frequency fingerprint synchronous identification model to obtain a link identifier corresponding to each sub-signal; wherein the multi-device radio frequency fingerprint synchronous identification model is trained based on the training method of the multi-device radio frequency fingerprint synchronous identification model provided by any embodiment of the present application; an identification module 804, used to perform multi-device radio frequency fingerprint synchronous identification based on the target link identifier and the mapping relationship between the pre-acquired link identifier and the device identifier.

[0117] Through the apparatus of the embodiment of the present application, accurate multi-device RF fingerprint synchronous identification can be performed in a multi-user multi-antenna scenario based on the trained multi-device RF fingerprint synchronous identification model.

[0118] It should be noted that the above explanation of the embodiment of the multi-device radio frequency fingerprint synchronous identification method is also applicable to the multi-device radio frequency fingerprint synchronous identification apparatus of this embodiment, and will not be repeated here.

[0119] In order to implement the above embodiment, the present application also proposes an electronic device. Figure 9 , Figure 9 Schematic diagram of the structure of the electronic device provided in the embodiment of the present application. Figure 9 As shown, the electronic device 900 includes: a processor 901, and a memory 902 communicatively connected to the processor 901; the memory 902 stores computer-executable instructions; the processor 901 executes the computer-executable instructions stored in the memory to implement the method provided in the aforementioned embodiment.

[0120] In order to implement the above embodiments, the present application also proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the methods provided by the above embodiments.

[0121] In order to implement the above embodiments, the present application also proposes a computer program product, including a computer program, which implements the methods provided by the above embodiments when executed by a processor.

[0122] In the description of this application, unless otherwise specified, " / " means or, for example, A / B can mean A or B; "and / or" in this article is only a way to describe the association relationship of associated objects, indicating that there can be three relationships, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0123] In the descriptions of the foregoing embodiments, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are mutually inconsistent.

[0124] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. Throughout the description of this application, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0125] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0126] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and a portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing it in another suitable manner if necessary, and then storing it in a computer memory.

[0127] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0128] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0129] In addition, the functional units in the various embodiments of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0130] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application. Persons skilled in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A training method for a multi-device radio frequency fingerprint synchronous recognition model, characterized in that: The multi-device radio frequency fingerprint synchronous identification model includes a first sub-model and a second sub-model, and the method includes: Acquire training data; wherein the training data includes a plurality of training signals and link identifiers of links corresponding to the plurality of training signals; performing signal preprocessing and multipath enhancement on the multiple training signals to obtain multiple sample signals; Inputting the multiple sample signals into a first initial sub-model to obtain multiple first tensors; wherein the first initial sub-model is used to process the multiple sample signals to obtain multiple tensors, and perform tensor slicing on the multiple tensors according to the number of links corresponding to the multiple signals to obtain the multiple first tensors; Input each of the first tensors into a second initial sub-model respectively to obtain a second tensor output by the second initial sub-model; A loss function is obtained based on the second tensor and the link identifier, and parameters of the first initial sub-model and the second initial sub-model are adjusted based on the loss function to obtain the first sub-model and the second sub-model.

2. The method according to claim 1, wherein The performing signal preprocessing and multipath enhancement on the multiple training signals to obtain multiple sample signals includes: Performing time-domain to frequency-domain transformation and normalization processing on the multiple training signals to obtain corresponding multiple first signals; Multipath channel enhancement and noise enhancement are performed on the multiple first signals to obtain the multiple sample signals.

3. The method according to claim 2, wherein Multipath channel enhancement is performed on the multiple first signals based on the following formula: Among them, s jl is the first signal at the jth link and the lth symbol before multipath channel enhancement, s ′ kl is the signal at the kth receiving link and the lth symbol after multipath channel enhancement, ° represents the element-by-element product, is the frequency domain response of a randomly generated multipath fading channel, N car is the number of subcarriers or frequencies.

4. The method according to claim 1, wherein The calculation formula of the loss function is: in, is the loss function, N tx is the number of links, r j is the jth second tensor, Represents r j c j The elements at position c j For r j The corresponding link identifier.

5. The method according to claim 1, wherein Both the first initial sub-model and the second initial sub-model include a residual module, the residual module includes at least one convolutional layer, and each convolutional layer includes a batch normalization layer and a Sigmoid activation layer.

6. A method for simultaneous identification of radio frequency fingerprints of multiple devices, characterized in that: include: Acquire a signal to be identified; wherein the signal to be identified includes multiple sub-signals; Performing signal preprocessing and multipath enhancement on the signal to be identified to obtain an input signal; Inputting the input signal into a multi-device radio frequency fingerprint synchronous identification model to obtain a target link identifier corresponding to each of the sub-signals; wherein the multi-device radio frequency fingerprint synchronous identification model is trained based on the method described in any one of claims 1 to 5; Based on the target link identifier and the mapping relationship between the pre-acquired link identifier and the device identifier, multi-device radio frequency fingerprint synchronous identification is performed.

7. A training device for a multi-device radio frequency fingerprint synchronous recognition model, characterized in that: The multi-device radio frequency fingerprint synchronous identification model includes a first sub-model and a second sub-model, and the apparatus includes: An acquisition module, configured to acquire training data; wherein the training data includes a plurality of training signals and link identifiers of links corresponding to the plurality of training signals; A first processing module is configured to perform signal preprocessing and multipath enhancement on the multiple training signals to obtain multiple sample signals; a second processing module, configured to input the plurality of sample signals into a first initial sub-model to obtain a plurality of first tensors; wherein the first initial sub-model is configured to process the plurality of sample signals to obtain a plurality of tensors, and to perform tensor slicing on the plurality of tensors according to the number of links corresponding to the plurality of signals to obtain the plurality of first tensors; a third processing module, configured to input each of the first tensors into a second initial sub-model to obtain a second tensor output by the second initial sub-model; The fourth processing module is used to obtain a loss function based on the second tensor and the link identifier, and adjust the parameters of the first initial sub-model and the second initial sub-model based on the loss function to obtain the first sub-model and the second sub-model.

8. A multi-device radio frequency fingerprint synchronous identification device, characterized in that: include: An acquisition module, configured to acquire a signal to be identified; wherein the signal to be identified includes a plurality of sub-signals; A first processing module is configured to perform signal preprocessing and multipath enhancement on the signal to be identified to obtain an input signal; A second processing module is configured to input the input signal into a multi-device radio frequency fingerprint synchronous recognition model to obtain a link identifier corresponding to each of the sub-signals; wherein the multi-device radio frequency fingerprint synchronous recognition model is trained based on the method of any one of claims 1 to 5; The identification module is used to perform simultaneous identification of radio frequency fingerprints of multiple devices based on the target link identifier and the mapping relationship between the pre-acquired link identifier and the device identifier.

9. An electronic device, characterized in that: include: a processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 5, or the method according to claim 6.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 5, or the method according to claim 6 when executed by a processor.

11. A computer program product, characterized in that The method comprises a computer program, which implements the method according to any one of claims 1 to 5, or the method according to claim 6 when executed by a processor.

Images