A radio frequency fingerprint extraction method and device based on a synchronization signal and a computer device

Abstract

This invention discloses a method, apparatus, and computer device for radio frequency fingerprint extraction based on synchronization signals. The method includes: acquiring multiple synchronization signal blocks, extracting a primary synchronization signal and / or a secondary synchronization signal to form a received signal set; performing channel estimation on each received signal in the received signal set and removing channel influence to obtain a signal set after preliminary removal of channel influence; grouping the signals after preliminary removal of channel influence in chronological order, and performing weighted superposition and denoising on multiple signals in each group to obtain a superimposed signal set; for the superimposed signal, utilizing the differences in channel characteristics at different times to remove channel differential-mode components, and further utilizing the channel consistency of the synchronization signals to remove channel common-mode components, thereby obtaining the radio frequency fingerprint. This invention can quickly obtain stable fingerprints before a terminal accesses a base station to identify illegal base stations, exhibits strong robustness, and effectively solves the problem of poor stability of radio frequency fingerprints in scenarios with rapidly changing channels, such as high-speed mobile environments.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication security, and relates to the extraction of radio frequency fingerprints from physical layer signals, and particularly to a method, apparatus and computer device for extracting radio frequency fingerprints based on synchronization signals. Background Technology

[0002] As core nodes in wireless communication networks, base stations' legitimacy and communication security directly impact terminal data transmission security and user privacy protection. In scenarios such as public network communication and private network deployment, illegal base stations can use methods like forging synchronization signals and broadcast information from legitimate base stations to lure terminals into accessing the network and engaging in malicious activities such as eavesdropping, fraud, and data tampering, posing a serious threat to mobile communication security. Therefore, achieving rapid and reliable identification of base station identity before a terminal connects to the base station has become a key technological requirement in the field of mobile communication security.

[0003] Radio frequency (RF) fingerprint extraction technology is a core method for base station identification. Its core principle is to utilize the inherent nonlinear characteristics, phase shift, amplitude distortion, and other physical features of the radio frequency front-end hardware (such as power amplifiers, filters, and mixers) of communication equipment to generate unique and stable RF fingerprints. Device identification is then achieved through fingerprint matching. Compared to traditional base station authentication methods based on higher-layer signaling, RF fingerprint extraction technology, based on physical layer signals, offers advantages such as faster authentication speed, stronger anti-forgery capabilities, and no reliance on higher-layer protocol interactions. It is particularly suitable for rapid identity verification scenarios before terminal access.

[0004] However, base station RF fingerprint extraction technology still faces many challenges. In wireless communication, signals are easily affected by the wireless channel environment during transmission, especially in scenarios with rapidly changing channels such as high-speed movement and dense urban areas. Interference such as multipath fading, Doppler frequency offset, and Gaussian white noise can mask the inherent distortion characteristics of the base station's RF front-end, resulting in poor stability and insufficient robustness of the extracted RF fingerprint. Patent application CN119364371A provides an RF fingerprint extraction method utilizing the logarithmic differential characteristics of the channel at different times; however, this method is only effective when there are significant differences between different signal channels and has poor adaptability to broadband signals. To achieve base station authentication before terminal access, there is an urgent need for an RF fingerprint extraction method that can quickly extract the base station's RF fingerprint before terminal access, while effectively dealing with complex interference in scenarios with rapidly changing channels, providing reliable authentication support for mobile communication security. Summary of the Invention

[0005] The purpose of this invention is to provide a radio frequency fingerprint extraction method, apparatus, and computer equipment based on synchronization signals, so as to solve the problems of existing radio frequency fingerprint extraction methods being greatly affected by channel interference, having poor adaptability to broadband signals, and being difficult to quickly identify before terminal access.

[0006] To achieve the above-mentioned objectives, in a first aspect, the present invention provides a radio frequency fingerprint extraction method based on a synchronization signal, comprising:

[0007] Collect multiple synchronization signal blocks, and extract the main synchronization signal and / or auxiliary synchronization signal to form a set of received signals;

[0008] Channel estimation is performed on each received signal in the received signal set to obtain the channel estimation result. Each received signal is then divided by the channel estimation result to obtain the signal set after preliminary removal of channel effects.

[0009] The signals after initial removal of channel effects are grouped in chronological order, and multiple signals in each group are weighted and superimposed for noise reduction to obtain a set of superimposed signals.

[0010] For the superimposed signals in the superimposed signal set, the channel differential mode components between the superimposed signals are removed by utilizing the channel characteristic differences at different times, and the calculation results containing only radio frequency characteristics and channel common mode components are obtained.

[0011] The calculation results, which include radio frequency characteristics and channel common-mode components, are used to remove the channel common-mode components by utilizing the channel consistency of the synchronization signal, thus obtaining the radio frequency fingerprint.

[0012] Further, the extraction of the primary synchronization signal and / or secondary synchronization signal includes: performing power normalization, timing synchronization calibration, carrier frequency offset compensation, cyclic prefix removal, and fast Fourier transform on each synchronization signal block signal in sequence, and extracting the primary synchronization signal and / or secondary synchronization signal from the transformed frequency domain signal.

[0013] Further, the process of performing channel estimation on each signal in the received signal set to obtain the channel estimation result includes:

[0014] Perform initial channel estimation on each received signal in the received signal set to obtain the initial channel estimate value;

[0015] The initial channel estimate is filtered and smoothed to obtain a smoothed channel estimate, which is used as the channel estimation result.

[0016] Furthermore, in the weighted superposition of multiple signals within each group, the weight of each signal is equal to the linear signal-to-interference-plus-noise ratio (SNR) of the corresponding signal divided by the sum of the SNRs of all signals within the group.

[0017] Furthermore, the step of utilizing the differences in channel characteristics at different times to remove the channel differential-mode components between the superimposed signals includes:

[0018] The superimposed signals in the superimposed signal set are paired up in pairs, and the channel differential mode component is removed by differential mode removal operation. For any two input signals, the differential mode removal operation is performed by calculating the ratio of the difference between the two input signals to the difference of the natural logarithm, and then obtaining the result after eliminating the channel differential mode component between the two input signals.

[0019] Furthermore, the step of utilizing the differences in channel characteristics at different times to remove the channel differential-mode components between the superimposed signals includes:

[0020] The superimposed signals in the superimposed signal set are paired up in pairs, and the channel differential mode component is removed by differential mode removal operation to obtain the first layer result; the differential mode removal operation is performed on any two input signals by calculating the ratio of the difference between the two input signals to the logarithmic difference to obtain the result after eliminating the channel differential mode component between the two input signals.

[0021] The first layer results are paired again, and the second layer results are obtained through differential mode removal. This process is repeated for multiple layers until the preset number of layers is reached, resulting in the channel differential mode components between the signals after desuperposition.

[0022] Furthermore, the removal of common-mode components from the channel using the channel consistency of the synchronization signal includes at least one of the following operations:

[0023] The calculation results of the main synchronization signal are used to perform a division operation between adjacent subcarriers based on the channel characteristic correlation of adjacent subcarriers;

[0024] The calculation results of the auxiliary synchronization signal are used to perform a division operation between adjacent subcarriers based on the channel characteristic correlation of adjacent subcarriers;

[0025] The calculation results of the primary synchronization signal and the calculation results of the secondary synchronization signal are divided by subcarrier.

[0026] Furthermore, at least one of the results of the adjacent subcarrier division operation corresponding to the synchronization signal or the auxiliary synchronization signal and the results of the subcarrier-by-subcarrier division operation is used as the extracted radio frequency fingerprint.

[0027] Secondly, the present invention provides a radio frequency fingerprint extraction device based on a synchronization signal, comprising:

[0028] The signal acquisition module is used to acquire multiple synchronization signal blocks and extract the main synchronization signal and / or auxiliary synchronization signal to form a received signal set;

[0029] The preliminary channel effect removal module is used to perform channel estimation on each received signal in the received signal set, obtain the channel estimation result, and divide each received signal by the channel estimation result to obtain the signal set after preliminary removal of channel effects;

[0030] The superposition denoising module is used to group the signals after the initial removal of channel effects into time order, and perform weighted superposition denoising on multiple signals in each group to obtain a superimposed signal set;

[0031] The channel differential mode component removal module is used to remove the channel differential mode components between the superimposed signals in the superimposed signal set by taking advantage of the channel characteristic differences at different times, and obtain the calculation results containing only radio frequency characteristics and channel common mode components.

[0032] The channel common-mode component removal module is used to remove the channel common-mode component from the calculation results containing radio frequency features and channel common-mode components by utilizing the channel consistency of the synchronization signal, so as to obtain the radio frequency fingerprint.

[0033] Thirdly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the radio frequency fingerprint extraction method described in the first aspect.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] 1. The radio frequency fingerprint extraction method provided by the present invention extracts radio frequency fingerprints by collecting multiple synchronization signal blocks and performing a series of signal processing steps. It can be used for rapid fingerprint extraction of the base station by the terminal before accessing the base station, so that the terminal can complete the base station identity recognition based on physical layer signals before accessing the base station, effectively preventing illegal base station forgery attacks and ensuring communication security from the physical layer.

[0036] 2. The radio frequency fingerprint extraction method provided by the present invention uses a synchronization signal for fingerprint extraction, which does not require high-level protocol interaction. It only requires signal processing at the receiving end, has a small computational load, and can quickly obtain fingerprint results, thus meeting the real-time requirements of rapid identity authentication before access.

[0037] 3. The RF fingerprint extraction method provided by this invention eliminates most of the channel influence through channel estimation and channel removal. Then, it improves the signal-to-noise ratio (SNR) through superposition denoising. Furthermore, it removes the differential-mode component of the channel by utilizing the differences in channel characteristics between signals at different times, and finally removes the common-mode component of the channel by utilizing the channel consistency of the synchronization signal. This ensures that the final RF fingerprint is only related to the inherent characteristics of the RF hardware, exhibiting high discriminability and stability. Compared with existing solutions, this invention effectively weakens the influence of channel factors on the signal during fingerprint extraction, making the extracted RF fingerprint more robust to multipath fading and channel noise, and insensitive to changes in the signal-to-noise ratio. This overcomes the insufficient applicability of existing technologies in broadband signal processing and maintains stable performance even in mobile environments and under conditions of rapid channel changes.

[0038] 4. The apparatus, electronic equipment and method provided by this invention have the same effect. Attached Figure Description

[0039] Figure 1 This is a schematic flowchart of a radio frequency fingerprint extraction method based on a synchronization signal provided in an embodiment of the present invention.

[0040] Figure 2 This is an example of the OFDM time-frequency structure diagram of 5G NR SSB in an embodiment of the present invention.

[0041] Figure 3 This is an example of the 5G NR SSB time-domain signal amplitude and cyclic prefix (CP) diagram in an embodiment of the present invention.

[0042] Figure 4 This is a frequency domain amplitude and phase diagram of the normalized signal from the commercial base station PSS in an embodiment of the present invention.

[0043] Figure 5 This is a diagram showing the channel estimation results for PSS and SSS in an embodiment of the present invention.

[0044] Figure 6 This is an amplitude diagram of PSS and SSS after preliminary channel removal in an embodiment of the present invention.

[0045] Figure 7 In this embodiment of the invention, the signal amplitude and residual diagrams after channel superposition of PSS and SSS at different times for two groups of the same base station are compared.

[0046] Figure 8 The amplitude and residual diagrams are calculated using the method of different channel characteristics at different times to compare the superposition results of two groups of signals from the same base station at different times in this embodiment of the invention.

[0047] Figure 9 These are two sets of fingerprint amplitude images extracted from the same base station at different times in this embodiment of the invention. Each set contains three fingerprint components.

[0048] Figure 10 The above are two sets of fingerprint amplitude maps extracted from different base stations at the same time in this embodiment of the invention. Each set contains three fingerprint components.

[0049] Figure 11 This is a schematic diagram of a radio frequency fingerprint extraction device based on a synchronization signal, provided in an embodiment of the present invention.

[0050] Figure 12 This is a schematic diagram of a computer device structure provided in an embodiment of the present invention. Detailed Implementation

[0051] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0052] like Figure 1 As shown in the figure, the radio frequency fingerprint extraction method based on synchronization signal provided by the embodiment of the present invention may include the following steps S1 to S5.

[0053] S1. Signal Acquisition: Acquire multiple synchronization signal blocks, extract the main synchronization signal and / or auxiliary synchronization signal to form a received signal set;

[0054] S2. Preliminary channel effect removal: Perform channel estimation on each received signal in the received signal set to obtain the channel estimation result, and divide each received signal by the channel estimation result to obtain the signal set after preliminary removal of channel effects;

[0055] S3. Superposition Denoising: The signals after the initial removal of channel influence are grouped in time order, and multiple signals in each group are weighted and superimposed for denoising to obtain a superimposed signal set.

[0056] S4. Channel differential mode component removal: For the superimposed signals in the superimposed signal set, the channel differential mode component between the superimposed signals is removed by utilizing the channel characteristic differences at different times, and the calculation result containing only radio frequency characteristics and channel common mode components is obtained.

[0057] S5. Channel common-mode component removal: The common-mode component is removed from the calculation results containing radio frequency features and channel common-mode components by utilizing the channel consistency of the synchronization signal to obtain the radio frequency fingerprint.

[0058] In this embodiment, steps S1 to S5 described above enable the extraction of the base station's radio frequency fingerprint before terminal access. By acquiring multiple synchronization signal blocks and extracting the radio frequency fingerprint through a series of signal processing steps, this can be used for rapid fingerprint extraction of the base station before terminal access. This allows the terminal to complete base station identification based on physical layer signals before access, effectively preventing unauthorized base station spoofing attacks and ensuring communication security at the physical layer. Furthermore, fingerprint extraction using synchronization signals eliminates the need for high-level protocol interaction; signal processing is only required at the receiving end, resulting in low computational load and rapid fingerprint acquisition, meeting the real-time requirements of rapid identity authentication before access. This embodiment eliminates most channel influences through channel estimation and removal, improves the signal-to-noise ratio through superposition denoising, removes channel differential-mode components by utilizing the channel characteristic differences between signals at different times, and finally removes channel common-mode components by utilizing the channel consistency of synchronization signals. This ensures that the final radio frequency fingerprint is only related to the inherent characteristics of the radio frequency hardware, exhibiting high discriminability and stability. Compared with existing solutions, this invention effectively weakens the influence of channel factors on the signal during fingerprint extraction, making the extracted radio frequency fingerprint more robust to multipath fading and channel noise, and insensitive to changes in signal-to-noise ratio. This overcomes the problem of insufficient applicability of existing technologies in broadband signal processing, and can still maintain stable performance in mobile environments and under conditions of rapid channel changes.

[0059] In some possible implementations, signal acquisition in step S1 may include the following steps S11 to S12.

[0060] S11. Signal Acquisition: In this embodiment, a 5G NR base station is selected for example. The 5G NR base station periodically transmits synchronization signal blocks (SSBs) according to the 3GPPTS 38.211 standard, such as... Figure 2 As shown, the SSB occupies 20 consecutive resource blocks (RBs) and 240 subcarriers in the frequency domain, including a primary synchronization signal (PSS), a secondary synchronization signal (SSS), and a physical broadcast channel (PBCH). The receiver performs time and frequency synchronization at the operating frequency of the base station and continuously receives at least K SSB signals, where K is an integer. In this embodiment, K=80, and the signal-to-interference ratio (SIR) of the received signals is filtered, retaining only signals with an SIR ≥ 10dB.

[0061] S12. Signal preprocessing: The received SSB signal is preprocessed sequentially. The preprocessing includes power normalization, timing synchronization calibration, carrier frequency offset (CFO) compensation, cyclic prefix (CP) removal, and fast Fourier transform (FFT).

[0062] In this embodiment, the timing synchronization calibration employs a synchronization algorithm based on PSS sequence correlation: leveraging the characteristic that the PSS sequence transmitted by the base station is a fixed sequence, a sliding correlation operation is performed between the PSS candidate interval in the received signal and the locally generated PSS standard sequence. By searching for the position of the correlation peak, the time-domain start position of the SSB signal and the Orthogonal Frequency Division Multiplexing (OFDM) symbol boundary are determined, achieving timing synchronization calibration with a synchronization accuracy ≤ 1 sampling point. CFO compensation employs a frequency offset estimation and removal method based on multiple CPs: extracting the cyclic prefix and corresponding repeated data segments at the tail of multiple consecutive OFDM symbols, estimating the carrier frequency offset of the received signal by calculating the phase difference of CPs between different symbols; then, CFO compensation and removal are performed on the time-domain received signal; finally, the CP segments in the original signal are removed. Figure 3 The amplitude of the 5G base station SSB time-domain signal received by the receiver is shown, and the CP segment of each OFDM symbol is marked.

[0063] In this embodiment, the 5G base station uses frequency band n41, the receiver sampling rate is 30.72MHz, and the FFT length of the OFDM demodulation process is... Both PSS and SSS are fixed sequences of length 127 in the frequency domain. , These correspond to the 127 subcarriers specified in the standard, with subcarrier spacing... The i-th signal received by the receiver , From the kth point of a fixed sequence , (hereinafter collectively referred to as) ) experienced base station RF front-end hardware RF distortion operator The transfer function is , Channel and noise , The synchronization signal is obtained later. The preprocessed synchronization signal satisfies:

[0064]

[0065]

[0066] The receiver receives the signal after power normalization. , The amplitude and phase are as follows Figure 4 As shown.

[0067] After preprocessing, step S1 obtains the received signal set. .

[0068] In this embodiment, the above preprocessing can accurately recover the frequency domain synchronization signal, eliminate timing and frequency offset errors at the receiving end, provide high-quality input data for subsequent processing, and thus improve the accuracy of fingerprint extraction.

[0069] In some possible implementations, the preliminary channel effect removal in step S2 may include the following steps S21 to S23.

[0070] S21. Initial Channel Estimation: Perform initial channel estimation on each received signal in the received signal set to obtain the initial channel estimation value;

[0071] S22. Filtering and smoothing: The initial channel estimate is filtered and smoothed to obtain a smoothed channel estimate, which is used as the channel estimation result.

[0072] S23. Divide the original received signal by the smoothed channel estimate to obtain the signal after preliminary removal of channel effects.

[0073] For ease of description, the main synchronization signal in step S1 will be referred to as... Auxiliary synchronization signal Merging is called Perform channel estimation and remove channel effects to obtain After step S2, the received signals in set S are processed to obtain a signal set that has been preliminarily freed from channel interference. .

[0074] For example, in this embodiment, least squares (LS) channel estimation is used: Let and to Filtering and smoothing are performed to obtain Finally, calculate and The quotient is obtained after the initial removal of channel effects. .

[0075] In this embodiment, the root mean square (RMS) delay is used to estimate the coherent bandwidth, thereby calculating the filter width. For a time-domain length of... The sampling points, let The absolute time delay corresponding to the nth time-domain sampling point. For time-domain channel impulse response, The power delay spectrum is in the time domain. First, the average arrival time of the channel impulse response is calculated. That is, the center position of the multipath signal power on the time delay axis, as shown in the following formula:

[0076]

[0077] Further calculation of root mean square delay spread:

[0078]

[0079] pass Calculate coherent bandwidth Filter smoothing window length ,in This is an empirical amplification factor, used in this embodiment. Calculate using filtering smoothing. In this embodiment, For parameters Normalized Gaussian kernel function: The channel estimation results are as follows: Figure 5 As shown. Final calculation. The results of removing channel effects are obtained, and after preliminary removal of channel effects, the following conditions are met:

[0080]

[0081] in For the residual channel components, It is residual Gaussian white noise. Figure 6 Showing Figure 5 The amplitude of the signal after removing channel effects is shown, such as Figure 6 As shown, after preliminary channel estimation, the signal amplitude fluctuates around 1, but due to noise, the fluctuation amplitude is relatively large.

[0082] In this embodiment, through the above-described preliminary channel estimation, filtering and smoothing, and channel removal processes, the channel effects in each received signal can be specifically eliminated, thus making the method still highly applicable in environments with rapidly changing channels.

[0083] In some possible implementations, in step S3, where multiple signals within each group are weighted and superimposed, the weight of each signal is equal to its linear signal-to-interference-plus-noise ratio (SNR) divided by the sum of the linear SNRs of all signals in its group. By weighting and superimposing the SNRs, signals with higher SNRs contribute more significantly, further optimizing the superimposed SNR and enhancing the denoising effect, making it particularly suitable for scenarios with fluctuating SNRs.

[0084] For example, step S3 may include steps S31 to S32 as follows.

[0085] S31. Signal Grouping: Group the signals after initial removal of channel effects in step S2. The signals in the data are divided into N non-overlapping groups according to time sequence, and each group... There are _ ...

[0086]

[0087] S32. Weighted Superposition: Weights are applied to the M signals in the j-th group and then summed to obtain the superimposed signal of that group. :

[0088]

[0089] in For the first The signal weights are superimposed. In this embodiment, the superimposed weights use:

[0090]

[0091] in For the first The linear signal-to-interference-plus-noise ratio (SIR) of a signal can be calculated by matching the received signal with a preset fixed sequence to obtain a signal power estimate, and then combining the noise and interference power to calculate the SIR.

[0092] Superimposed signal satisfy:

[0093]

[0094] in and Residual channel components and residual noise Formed by stacking in the same way.

[0095] After step S3, the superimposed signal set is obtained. .

[0096] In this embodiment, after initial superposition, the signal-to-interference-plus-noise ratio (SIR) is greater than 20 dB, suppressing Gaussian white noise. . Figure 7 Comparing the amplitude and residual of the signal after superimposing two PSSs (Pulse Signal Spectra) at different times from the same base station, it can be seen that the signal amplitude stabilizes around 1 after superposition, the residual amplitude of the two signals is small, and the noise is suppressed to a certain extent; residual channel components... satisfy ,in .

[0097] In some possible implementations, the step S4, which utilizes the channel characteristic differences at different times to remove the channel differential mode components between the superimposed signals, may include the following steps S41 to S42.

[0098] S41. Define the differential mode removal operation: For any two input signals, define a differential mode removal operation function whose output is equal to the difference between the two input signals divided by the difference between their natural logarithms.

[0099] S42, Removal Calculation: Pair the superimposed signals in the superimposed signal set into pairs and remove the channel differential mode components through differential mode removal operation.

[0100] In this embodiment, the differential mode removal operation can eliminate the channel differential mode component between two signals while retaining their common mode component.

[0101] Preferably, the calculation result of step S42 is used as the first-level result, and further includes:

[0102] S43, Multi-level Progression: The first-level results are paired up again, and the second-level results are obtained through differential mode removal operation. This process is repeated for multiple levels until the preset number of levels is reached. The channel differential mode components between the superimposed signals are removed to obtain the common mode components between all superimposed signals.

[0103] This embodiment employs multi-layered progressive calculations, leveraging the differences in channel characteristics among multiple signals in the set to gradually eliminate differential-mode components, thereby extracting common-mode features among the signals. This layered suppression of channel influences in the signal, combined with processing of channel characteristic differences between signals at multiple time points, makes the extracted RF fingerprint robust to multipath fading and channel noise, insensitive to changes in signal-to-noise ratio, and particularly effective in scenarios with rapidly changing channels, such as mobile scenarios and environmental changes.

[0104] More specifically, for the superimposed signal obtained in step S3 This paper describes a method based on the differences in channel characteristics at different times to remove the differential mode components of the channel at different times under the condition of rapid channel change.

[0105] Through step S3, we have ,make:

[0106]

[0107] Further have:

[0108]

[0109] in The residual channel components in step S3, It includes the residual noise from step S3. for and The merged items include effects other than radio frequency fingerprinting.

[0110] For the same base station, take two different times or environments. :

[0111]

[0112]

[0113] In this case, the two environments may be composed of the same channel components. And the differential-mode component caused by residual channel or noise. , composition.

[0114] make:

[0115]

[0116] but:

[0117]

[0118] According to step S3, the residual channel components satisfy ,in Residual noise ;make , satisfy Substitute There is:

[0119]

[0120]

[0121] So:

[0122]

[0123] Will The Taylor expansion around 1 is performed while preserving the first-order approximation:

[0124]

[0125] Substituting the above equation into... There is:

[0126]

[0127] Define the modulus removal operation function F:

[0128]

[0129] From the above, it is easy to conclude that:

[0130]

[0131] At this point, Only radio frequency fingerprint features and , Common-mode components of the residual channel .

[0132] Furthermore, for sets Select as well as The two pairs of distinct signals are calculated using the method described above:

[0133]

[0134]

[0135] in and Each retained the channel common-mode component. , ; Channel common-mode components , Further represented as common-mode components and differential mode components , .Compare and The expression and and The expressions can be found to be of the same form, therefore further analysis is possible. and Perform calculations to remove their differential modulus components:

[0136]

[0137] Following this logic, we obtain the multi-level progressive calculation formula:

[0138]

[0139] in This represents the result of removing the differential mode component of the channel at layer L. (Subscript) It is only used to distinguish between two different signals selected in each layer and to differentiate between different layers. Through multi-layer progressive calculation, it is possible to remove the channel differential-mode components from a set of datasets step by step, while retaining the common-mode components.

[0140] As an optional implementation, for the result obtained in step S3 Based on the chronological order, the channel differential mode components are removed step by step. Specifically:

[0141]

[0142] in , For the first The total number of signals in the layer. First, for , will the One signal With the One signal The differential removal function F is used for calculation to obtain the initial layer results. Then, recursive calculations are performed. For the Lth layer, , , will the first of the upper layer Results With the Results The differential removal function F is used to calculate the result of the current layer. In this embodiment, the final layer is selected. In the final layer The final output contains only the calculation results of radio frequency characteristics and channel common-mode components. satisfy:

[0143]

[0144] in For set The common-mode components of all signals in the set can be compared to the greatest common divisor (GCD) of the set.

[0145] Figure 8 By comparing the results of two sets of PSS at different times from the same base station and the SSS after superposition using the method described above, it can be seen that the two sets of signals processed by this method tend to be consistent, and the residual is further reduced.

[0146] In some possible implementations, the removal of the common-mode component of the channel in step S5 can be based on the consistency of the channel characteristics of different synchronization signals within the same synchronization signal block, performing a subcarrier-by-subcarrier complex operation on the extraction results of the main synchronization signal and the auxiliary synchronization signal, or performing a complex operation on adjacent subcarriers based on the correlation of the channel characteristics of adjacent subcarriers, or other equivalent processing strategies can be adopted.

[0147] Preferably, step S5, which utilizes the channel consistency of the synchronization signal to remove the common-mode component of the channel, includes at least one of the following operations: performing adjacent subcarrier division on the calculation result of the primary synchronization signal based on the channel characteristic correlation of adjacent subcarriers; performing adjacent subcarrier division on the calculation result of the secondary synchronization signal based on the channel characteristic correlation of adjacent subcarriers; and performing subcarrier-by-subcarrier division on the calculation result of the primary synchronization signal and the calculation result of the secondary synchronization signal. At least one of the results of the adjacent subcarrier division operation corresponding to the synchronization signal or the secondary synchronization signal, and the results of the subcarrier-by-subcarrier division operation, is used as the extracted radio frequency fingerprint.

[0148] In practical applications, one or more methods can be selected to remove the channel common-mode component, resulting in different fingerprint components. These fingerprint components can be used individually or in combination.

[0149] For example, based on step S4, it is known that The changed channel components have been removed:

[0150]

[0151]

[0152] Because the subcarrier spacing of the 5G NR synchronization signal (PSS / SSS) is small, the channel characteristics of adjacent subcarriers are approximately the same. , The adjacent subcarrier channels have , .right , Perform the division operation between adjacent subcarriers respectively:

[0153]

[0154]

[0155] Furthermore, based on step S1, it can be known that for two fixed sequences within the same frame... and Having experienced similar channels, there are Therefore, , Dividing them gives us:

[0156]

[0157] At this point, , , The influence of the channel has been effectively removed, and the issue is only related to two different ideal data signals and the transmitter's RF characteristics. In this embodiment, the combined processing results of the above three segments are... One or more combinations of these are considered as the radio frequency fingerprint of the base station. Radio frequency fingerprints extracted from the same base station at different times using the method described in this embodiment are as follows: Figure 9 As shown, the radio frequency fingerprints of the same base station at different times are relatively stable; the radio frequency fingerprints extracted from different base stations at the same time using the method described in this embodiment are as follows: Figure 10 As shown, the radio frequency fingerprints from different base stations at the same time have significant distinguishability.

[0158] The radio frequency fingerprint extraction device according to the embodiments of the present invention will be further described below. The radio frequency fingerprint extraction device described below can be referred to in correspondence with the radio frequency fingerprint extraction method described above.

[0159] like Figure 11 As shown in the figure, an embodiment of the present invention provides a radio frequency fingerprint extraction device based on synchronization signals, comprising: a signal acquisition module 11, used to acquire multiple synchronization signal blocks, extract the main synchronization signal and / or auxiliary synchronization signal to form a received signal set; a preliminary channel influence removal module 12, used to perform channel estimation on each received signal in the received signal set, obtain a channel estimation result, and divide each received signal by the channel estimation result to obtain a signal set after preliminary removal of channel influence; a superposition denoising module 13, used to group the signals after preliminary removal of channel influence in chronological order, perform weighted superposition denoising on multiple signals in each group, and obtain a superimposed signal set; a channel differential mode component removal module 14, used to remove the channel differential mode component between the superimposed signals in the superimposed signal set by utilizing the channel characteristic differences at different times, to obtain a calculation result containing only radio frequency features and channel common mode components; and a channel common mode component removal module 15, used to remove the channel common mode component from the calculation result containing radio frequency features and channel common mode components by utilizing the channel consistency of the synchronization signal, to obtain a radio frequency fingerprint.

[0160] like Figure 12 As shown, this embodiment of the invention also provides a computer device, including a memory 21, a processor 22, and a computer program stored on the memory 21 and executable on the processor 22. When the computer program is executed by the processor 22, it implements the steps of the radio frequency fingerprint extraction method based on synchronization signal in any of the above embodiments.

[0161] The program code used to implement the method of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the steps of the method of the present invention to be performed.

[0162] The radio frequency fingerprint extraction device and computer equipment provided in this embodiment have the same beneficial effects as the radio frequency fingerprint extraction method described above, and will not be repeated here.

[0163] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems or apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and relevant parts can be referred to the method section.

[0164] It should also be noted that relational terms such as "first" and "second" in this specification are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The term "and / or" includes any and all combinations of one or more of the associated listed items. The terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0165] It should be noted that the above-described embodiments only illustrate some implementation methods of the present invention, and their description should not be construed as limiting the scope of the present invention. It should be pointed out that those skilled in the art can make several improvements without departing from the concept of the present invention, and these improvements should all fall within the protection scope of the present invention.

Claims

1. A radio frequency fingerprint extraction method based on synchronization signals, characterized in that, include: Collect multiple synchronization signal blocks, and extract the main synchronization signal and / or auxiliary synchronization signal to form a set of received signals; Channel estimation is performed on each received signal in the received signal set to obtain the channel estimation result. Each received signal is then divided by the channel estimation result to obtain the signal set after preliminary removal of channel effects. The signals after initial removal of channel effects are grouped in chronological order, and multiple signals in each group are weighted and superimposed for noise reduction to obtain a set of superimposed signals. For the superimposed signals in the superimposed signal set, the channel differential mode component between the superimposed signals is removed by utilizing the channel characteristic differences at different times, and the calculation results including radio frequency characteristics and channel common mode components are obtained. The calculation results, which include radio frequency characteristics and channel common-mode components, are used to remove the channel common-mode components by utilizing the channel consistency of the synchronization signal, thus obtaining the radio frequency fingerprint.

2. The radio frequency fingerprint extraction method according to claim 1, characterized in that, The extraction of the primary synchronization signal and / or secondary synchronization signal includes: performing power normalization, timing synchronization calibration, carrier frequency offset compensation, cyclic prefix removal, and fast Fourier transform on each synchronization signal block signal in sequence, and extracting the primary synchronization signal and / or secondary synchronization signal from the transformed frequency domain signal.

3. The radio frequency fingerprint extraction method according to claim 1, characterized in that, The process of performing channel estimation on each signal in the received signal set to obtain the channel estimation result includes: Perform initial channel estimation on each received signal in the received signal set to obtain the initial channel estimate value; The initial channel estimate is filtered and smoothed to obtain a smoothed channel estimate, which is used as the channel estimation result.

4. The radio frequency fingerprint extraction method according to claim 1, characterized in that, In the weighted superposition of multiple signals within each group, the weight of each signal is equal to the linear signal-to-interference-plus-noise ratio (SNR) of the corresponding signal divided by the sum of the SNRs of all signals in the group.

5. The radio frequency fingerprint extraction method according to claim 1, characterized in that, The method of utilizing the differences in channel characteristics at different times to remove the channel differential mode components between superimposed signals includes: The superimposed signals in the superimposed signal set are paired up in pairs, and the channel differential mode component is removed by differential mode removal operation. For any two input signals, the differential mode removal operation is performed by calculating the ratio of the difference between the two input signals to the difference of the natural logarithm, and then obtaining the result after eliminating the channel differential mode component between the two input signals.

6. The radio frequency fingerprint extraction method according to claim 1, characterized in that, The method of utilizing the differences in channel characteristics at different times to remove the channel differential mode components between superimposed signals includes: The superimposed signals in the superimposed signal set are paired up in pairs, and the channel differential mode component is removed by differential mode removal operation to obtain the first layer result; the differential mode removal operation is performed on any two input signals by calculating the ratio of the difference between the two input signals to the logarithmic difference to obtain the result after eliminating the channel differential mode component between the two input signals. The first layer results are paired again, and the second layer results are obtained through differential mode removal. This process is repeated for multiple layers until the preset number of layers is reached, resulting in the channel differential mode components between the signals after desuperposition.

7. The radio frequency fingerprint extraction method according to claim 1, characterized in that, The method of removing common-mode components from the channel by utilizing the channel consistency of the synchronization signal includes at least one of the following operations: The calculation results of the main synchronization signal are used to perform a division operation between adjacent subcarriers based on the channel characteristic correlation of adjacent subcarriers; The calculation results of the auxiliary synchronization signal are used to perform a division operation between adjacent subcarriers based on the channel characteristic correlation of adjacent subcarriers; The calculation results of the primary synchronization signal and the calculation results of the secondary synchronization signal are divided by subcarrier.

8. The radio frequency fingerprint extraction method according to claim 7, characterized in that, The result of the division operation between adjacent subcarriers corresponding to the synchronization signal or the auxiliary synchronization signal and the result of the subcarrier-by-subcarrier division operation are used as the extracted radio frequency fingerprint.

9. A radio frequency fingerprint extraction device based on a synchronization signal, characterized in that, include: The signal acquisition module is used to acquire multiple synchronization signal blocks and extract the main synchronization signal and / or auxiliary synchronization signal to form a received signal set; The preliminary channel effect removal module is used to perform channel estimation on each received signal in the received signal set, obtain the channel estimation result, and divide each received signal by the channel estimation result to obtain the signal set after preliminary removal of channel effects; The superposition denoising module is used to group the signals after the initial removal of channel effects into time order, and perform weighted superposition denoising on multiple signals in each group to obtain a superimposed signal set; The channel differential mode component removal module is used to remove the channel differential mode components between the superimposed signals in the superimposed signal set by taking advantage of the channel characteristic differences at different times, and obtain the calculation results containing only radio frequency characteristics and channel common mode components. The channel common-mode component removal module is used to remove the channel common-mode component from the calculation results containing radio frequency features and channel common-mode components by utilizing the channel consistency of the synchronization signal, so as to obtain the radio frequency fingerprint.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is executed by the processor, it implements the radio frequency fingerprint extraction method according to any one of claims 1-8.

Images