Mobile phone geographic position detection system based on mobile communication signal

Abstract

The invention belongs to the technical field of mobile communication and mobile phone positioning, and particularly provides a mobile phone geographic position detection system based on a mobile communication signal, and the system comprises a data receiving module which is used for receiving physical layer channel impulse response waveform data and protocol layer signaling data from a mobile phone end; the data processing module is used for carrying out dynamic characteristic correction and screening on the physical layer channel impulse response waveform data to generate a frequency offset compensation waveform; analyzing the protocol layer signaling data, and generating protocol layer signaling anomaly statistics; and the frequency domain feature extraction module is used for performing wavelet packet decomposition on the frequency offset compensation waveform to generate an effective frequency band energy matrix. According to the method and the device, high-precision calculation of the geographic coordinates of the mobile phone terminal can be realized, the geographic coordinates of the mobile phone terminal can be accurately generated, the risk value of the pseudo base station can be given, the use safety of the mobile phone and the accuracy of position detection are improved, and the communication safety of a user and the reliability of position information acquisition are effectively guaranteed.

Description

Technical Field

[0001] The present disclosure belongs to the technical field of mobile communication and mobile phone positioning, and particularly relates to a mobile phone geographical location detection system based on mobile communication signals. Background Art

[0002] The analysis of the spatial propagation characteristics based on the multipath delay or the frequency-domain energy characteristics of the channel impulse response waveform is often used to estimate the geographical location of the signal source.

[0003] In a dynamic mobile scenario, the relative movement between the terminal and the base station causes Doppler frequency offset and environmental time-variation, resulting in non-linear distortion of the physical layer signal characteristics. Such distortion masks the inherent differences between the rogue base station and the legitimate base station in terms of hardware or deployment, such as abnormal propagation delay and frequency-domain energy offset. Traditional techniques rely on static feature weights and fixed classification thresholds, and it is difficult to dynamically model the correlation between the frequency offset characteristics and interference signals in a time-varying channel, resulting in inaccurate quantification of the environmental interference intensity. The positioning algorithm is difficult to correct the dynamic contamination of the propagation parameters by the rogue base station, leading to the continuous accumulation of deviations in the calculation of the geographical coordinates at the mobile phone end and prone to offset phenomena. Summary of the Invention

[0004] The present disclosure effectively solves the problem in the prior art that due to the static processing of dynamic channel interference, it is impossible to accurately model the time-varying correlation, resulting in the ineffective suppression of the rogue base station signal pollution and the continuous offset of the mobile phone positioning coordinates due to the distortion of dynamic parameters. The present disclosure dynamically adjusts the classification weights and thresholds, optimizes the spatio-temporal continuity of the channel labels, enables the credibility score to accurately quantify the interference intensity of the rogue base station, and finally fuses the abnormal statistics at the protocol layer to correct the distortion of the multipath delay parameters caused by the rogue base station signal, realizing the high-precision calculation of the geographical coordinates at the mobile phone end.

[0005] To achieve the above object, the present disclosure adopts the following technical solutions: In a first aspect, the present disclosure provides a mobile phone geographical location detection system based on mobile communication signals, including: A data receiving module, which is used to receive physical layer channel impulse response waveform data and protocol layer signaling data from the mobile phone side; a data processing module, which is used to perform dynamic feature correction and screening on the physical layer channel impulse response waveform data to generate a frequency offset compensation waveform; parse the protocol layer signaling data to generate protocol layer signaling anomaly statistics; a frequency domain feature extraction module, which is used to perform wavelet packet decomposition on the frequency offset compensation waveform to generate an effective band energy matrix; a pseudo base station identification module, which is used to input the generated effective band energy matrix into a pseudo base station identification model. The pseudo base station identification model is based on a random forest classifier, and adjusts the Doppler frequency offset feature weight and classification threshold according to the effective correlation period, and outputs a channel environment classification label; a comprehensive analysis module, which is used to perform a sliding window weighted average on the channel environment classification label to generate an environment credibility score; a risk assessment and positioning module, which is used to generate a pseudo base station risk value and the geographical coordinates of the mobile phone side according to the environment credibility score and the protocol layer signaling anomaly statistics.

[0006] Further, performing dynamic feature correction and screening on the physical layer channel impulse response waveform data to generate a frequency offset compensation waveform includes: calculating the dynamic parameters of the physical layer channel impulse response waveform to obtain the multipath delay spread and Doppler frequency offset; dynamically intercepting the main path part of the physical layer channel impulse response waveform according to the multipath delay spread to generate a delay correction waveform; performing phase compensation on the waveform according to the Doppler frequency offset to generate a frequency offset compensation waveform.

[0007] Further, calculate the time-varying correlation coefficient sequence of the delay correction waveform and the frequency offset compensation waveform; set a correlation threshold and a time threshold; screen the time period in the time-varying correlation coefficient sequence where the absolute value of the correlation coefficient is greater than the correlation threshold and the duration is greater than the time threshold to generate an effective correlation period mark; perform wavelet packet decomposition on the frequency offset compensation waveform within the interval of the effective correlation period.

[0008] Further, performing wavelet packet decomposition on the frequency offset compensation waveform to generate an effective band energy matrix includes: performing wavelet packet decomposition on the frequency offset compensation waveform to obtain sub-band coefficients of each layer; using a smooth soft threshold function to denoise the sub-band coefficients, and calculating the sub-band energy by weighting according to the importance of the sub-band in the mobile communication signal; extracting a specified number of sub-band energies to generate an effective band energy matrix.

[0009] Further, the protocol layer signaling anomaly statistics include the IMSI request frequency; adjusting the Doppler frequency offset feature weight according to the effective correlation period includes: setting a frequency threshold and a first proportional value; obtaining the sudden increase amount of the IMSI request frequency; when the sudden increase amount of the IMSI request frequency is greater than the frequency threshold, increasing the splitting gain weight of the Doppler frequency offset related sub-band to the first proportional value.

[0010] Furthermore, the protocol layer signaling anomaly statistics include the number of TA value jumps; adjusting the classification threshold according to the effective correlation period includes: setting a jump threshold and a second proportional value; when the number of TA value jumps is greater than the jump threshold, the splitting gain weight of the sub-bands related to the multipath delay spread is attenuated to the second proportional value.

[0011] Furthermore, the channel environment classification label includes at least Label 1 and Label 2; in the wavelet packet decomposition of the frequency offset compensation waveform, the number of wavelet packet decomposition layers is the first specified number of layers; setting a second specified number of layers, the second specified number of layers is greater than the first specified number of layers; after outputting the channel environment classification label: if the channel environment classification label is Label 1, then increase the number of wavelet packet decomposition layers in the next period from the first specified number of layers to the second specified number of layers; if the channel environment classification label is Label 2, then shorten the sliding window length to the specified window length.

[0012] Furthermore, the node splitting rule of the random forest classifier in the false base station identification model includes: calculating the protocol anomaly score; setting an anomaly threshold and a mean threshold; if the protocol anomaly score is greater than the anomaly threshold, then select the Doppler frequency offset related features for node splitting; if the protocol anomaly score is less than or equal to the anomaly threshold, then select the multipath delay spread related features for node splitting; among them, the selected related features satisfy that their historical splitting gain mean under the current protocol anomaly score is greater than the mean threshold.

[0013] Furthermore, generating the false base station risk value and the mobile phone end geographical coordinates according to the environmental credibility score and the protocol layer signaling anomaly statistics includes: inputting the environmental credibility score and the protocol layer signaling anomaly statistics into the spatio-temporal gating decision network, and outputting the false base station risk value and the geographical coordinates; performing distance compensation and secondary correction on the geographical coordinates to generate the mobile phone end coordinates; among them, the spatio-temporal gating decision network includes: Physical branch: Extracting the signal stability feature from the environmental credibility score through the first fully connected layer, setting the negative interval slope of the activation function to the specified proportion of the effective correlation period ratio, and generating a physical feature vector; Protocol branch: Extracting the timing feature from the number of TA jumps and the IMSI anomaly frequency through the gated recurrent unit network, and generating a protocol feature vector; Fusion layer: Dynamically adjusting the weights of the physical branch and the protocol branch according to the effective correlation period ratio, and generating a fusion feature; Second fully connected layer: Outputting the false base station risk value and the geographical coordinates of the mobile phone end according to the fusion feature.

[0014] Further, distance compensation and secondary correction are performed on the geographical coordinates to generate the mobile phone end coordinates, including: calculating the preliminary distance between the base station and the terminal according to the geographical coordinates; calculating the compensated distance according to the Doppler frequency offset and the preliminary distance; setting a proportion threshold; if the proportion of the effective association period is greater than or equal to the proportion threshold, using the coordinates after the compensated distance as the mobile phone end coordinates; if the proportion of the effective association period is less than the proportion threshold, performing secondary correction according to the Doppler frequency offset and the compensated distance, and using the secondary corrected geographical coordinates as the mobile phone end coordinates.

[0015] In a second aspect, the present disclosure provides a mobile phone geographical location detection device based on mobile communication signals, including: A radio frequency front-end module, configured to receive the physical layer radio frequency signal of the mobile phone end, and generate physical layer channel impulse response waveform data through down-conversion and analog-to-digital conversion.

[0016] A protocol parsing chip, connected to the radio frequency front-end module, configured to demodulate protocol layer signaling data from the physical layer channel impulse response waveform data, and count protocol layer signaling anomaly indicators.

[0017] An FPGA chip, connected to the radio frequency front-end module, configured to: perform dynamic feature correction and screening on the physical layer channel impulse response waveform data to generate a frequency offset compensated waveform; perform wavelet packet decomposition on the frequency offset compensated waveform to generate an effective band energy matrix.

[0018] A main control unit, connected to the FPGA chip and the protocol parsing chip, configured to: input the effective band energy matrix into a pseudo base station identification model, the pseudo base station identification model is based on a random forest classifier, dynamically adjust the Doppler frequency offset feature weight and classification threshold according to the effective association period, and output a channel environment classification label; perform a sliding window weighted average on the channel environment classification label to generate an environment credibility score; generate a pseudo base station risk value and mobile phone end geographical coordinates according to the environment credibility score and protocol layer signaling anomaly statistics.

[0019] Advantages of the present disclosure: By dynamically adjusting the classification weight and threshold, the present disclosure optimizes the spatio-temporal continuity of the channel label, enables the credibility score to accurately quantify the interference intensity of the pseudo base station, and finally fuses the protocol layer anomaly statistics to correct the distortion of the multi-path delay parameters caused by the pseudo base station signal, realizing high-precision calculation of the mobile phone end geographical coordinates, effectively solving the problems in the prior art that due to static processing of dynamic channel interference, the time-varying correlation cannot be accurately modeled, resulting in ineffective suppression of pseudo base station signal pollution and continuous deviation of the mobile phone positioning coordinates due to dynamic parameter distortion. It can not only accurately generate the mobile phone end geographical coordinates, but also give the pseudo base station risk value, improving the security of mobile phone use and the accuracy of location detection, and effectively ensuring the reliability of user communication security and location information acquisition.

[0020] Other features and advantages of the present disclosure will be set forth in the following description, and in part will be obvious from the description, or may be learned by practice of the present disclosure. The objectives and other advantages of the present disclosure may be realized and attained by the structure particularly pointed out in the specification and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] To more clearly illustrate the technical solutions in the embodiments of the present disclosure or in the prior art, the following briefly introduces the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are some embodiments of the present disclosure. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0022] Figure 1 FIG. shows a schematic diagram of a mobile phone geographical location detection system based on mobile communication signals according to the present disclosure; Figure 2 FIG. shows a schematic diagram of a mobile phone geographical location detection device based on mobile communication signals according to the present disclosure. DETAILED DESCRIPTION

[0023] To solve the problems raised in the background art, the present disclosure dynamically adjusts the classification weights and thresholds, optimizes the spatio-temporal continuity of channel tags, accurately quantifies the intensity of pseudo base station interference with credibility scores, and finally fuses the abnormal statistics at the protocol layer to correct the distortion of the pseudo base station signal on the multipath delay parameters, realizing high-precision calculation of the geographical coordinates at the mobile phone end.

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure clearer, the following clearly and completely describes the technical solutions in the embodiments of the present disclosure with reference to the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are some, but not all, of the embodiments of the present disclosure. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present disclosure without creative efforts fall within the scope of protection of the present disclosure.

[0025] In some embodiments, as Figure 1 shown, the present disclosure provides a mobile phone geographical location detection system based on mobile communication signals, including: a data receiving module, a data processing module, a frequency domain feature extraction module, a pseudo base station identification module, a comprehensive analysis module, and a risk assessment and positioning module.

[0026] The data receiving module is configured to receive the physical layer channel impulse response waveform data and the protocol layer signaling data from the mobile phone end.

[0027] The data processing module is used to perform dynamic feature correction and screening on the physical layer channel impulse response waveform data to generate a frequency offset compensation waveform; parse the protocol layer signaling data to generate protocol layer signaling anomaly statistics.

[0028] The frequency domain feature extraction module is used to perform wavelet packet decomposition on the frequency offset compensation waveform to generate an effective band energy matrix.

[0029] The fake base station identification module is used to input the generated effective band energy matrix into the fake base station identification model. The fake base station identification model is based on a random forest classifier and adjusts the Doppler frequency offset feature weight and classification threshold according to the effective correlation period, and outputs a channel environment classification label.

[0030] The comprehensive analysis module is used to perform sliding window weighted averaging on the channel environment classification label to generate an environment credibility score.

[0031] The risk assessment and positioning module is used to generate a fake base station risk value and the geographical coordinates of the mobile phone end according to the environment credibility score and the protocol layer signaling anomaly statistics.

[0032] In some embodiments, performing dynamic feature correction and screening on the physical layer channel impulse response waveform data to generate a frequency offset compensation waveform includes: calculating the dynamic parameters of the physical layer channel impulse response waveform to obtain the multipath delay spread and Doppler frequency offset; dynamically intercepting the main path part of the physical layer channel impulse response waveform according to the multipath delay spread to generate a delay correction waveform; performing phase compensation on the waveform according to the Doppler frequency offset to generate a frequency offset compensation waveform.

[0033] Based on the physical layer channel impulse response waveform data, the root mean square delay spread formula is used to calculate the multipath delay spread . The specific steps include: extracting the delay and corresponding power of each path in the waveform; calculating the multipath delay spread through the root mean square delay spread formula .

[0034] Exemplarily, in the LTE signal, if the detected delays of three paths are 0.2 μs, 0.8 μs, and 1.5 μs respectively, and the corresponding powers are 0.3, 0.5, and 0.2, the calculated multipath delay spread is 0.63 μs.

[0035] The phase difference method is used to calculate the Doppler frequency offset . The specific steps include: calculating the phase difference of adjacent signal sampling points; estimating the Doppler frequency offset according to the phase difference sequence .

[0036] Exemplarily, in the 5G NR system, if the sampling interval is 0.1 μs, the calculated Doppler frequency offset is 450 Hz, corresponding to a terminal moving speed of approximately 80 km / h.

[0037] According to the multipath delay spread , dynamically intercept the main path part of the physical layer channel impulse response waveform. The main path determination rule can be: the main path delay is the path delay with the maximum power, and the interception window range is , where m is a dynamically adjustable coefficient, and the value range is 0.2 to 0.5.

[0038] Exemplarily, if the multipath delay spread = 0.63 μs and m = 0.3, then the interception window is the main path delay ±0.19 μs.

[0039] Based on the multipath delay spread adaptive adjustment of the interception range can ensure the accuracy of the main path selection.

[0040] Perform phase compensation on the delay correction waveform to eliminate the influence of Doppler frequency offset. The phase compensation method can be: according to the Doppler frequency offset , apply a phase compensation amount to each sampling point of the delay correction waveform, where n is the sampling point number and T is the sampling interval.

[0041] Exemplarily, if the Doppler frequency offset is , and the sampling interval T is , then the phase compensation amount for each sampling point is ≈ -0.028 radians.

[0042] Perform real-time frequency offset compensation through the phase difference method, which can reduce the compensation delay.

[0043] In some embodiments, calculate the time-varying correlation coefficient sequence of the delay correction waveform and the frequency offset compensation waveform; set a correlation threshold and a time threshold; screen the time-varying correlation coefficient sequence to find the time period where the absolute value of the correlation coefficient is greater than the correlation threshold and the duration is greater than the time threshold, and generate a valid correlation period mark; within the interval of the valid correlation period, perform wavelet packet decomposition on the frequency offset compensation waveform.

[0044] Segment the delay correction waveform and the frequency offset compensation waveform with a fixed window length and step size. For example, the window length is 10 ms and the step size is 1 ms.

[0045] Calculate the correlation coefficient for the two waveform data within each window to generate a time-varying correlation coefficient sequence.

[0046] Exemplarily, in a 5G NR signal, if the window length is 10 ms and it contains 1,000 sampling points, the calculated correlation coefficient sequence is [-0.2, 0.85, 0.92, -0.1, 0.88].

[0047] The correlation threshold can be set to the absolute value of 0.8 for screening high-correlation time periods.

[0048] The time threshold can be set to 5 ms to avoid misjudgment caused by instantaneous interference.

[0049] Exemplarily, if the correlation coefficient sequence is [0.85, 0.92, 0.88] and it lasts for 3 windows, which is 3 ms and does not meet the time threshold; if the sequence is [0.85, 0.92, 0.88, 0.91, 0.89] and it lasts for 5 ms, it is marked as a valid correlation time period.

[0050] Only within the corresponding interval marked during the valid correlation time period, wavelet packet decomposition is performed on the frequency offset compensation waveform; data during non-valid time periods is temporarily stored or discarded, which can reduce invalid calculations.

[0051] In some embodiments, wavelet packet decomposition is performed on the frequency offset compensation waveform to generate an effective frequency band energy matrix, including: performing wavelet packet decomposition on the frequency offset compensation waveform to obtain sub-band coefficients of each layer; using a smooth soft threshold function to denoise the sub-band coefficients, and calculating the sub-band energy by weighting according to the importance of the sub-band in the mobile communication signal; extracting a specified number of sub-band energies to generate an effective frequency band energy matrix.

[0052] The wavelet basis function can be selected as the Daubechies wavelet; the initial number of layers is the first specified number of layers, such as 4 layers.

[0053] Performing full-tree decomposition on the frequency offset compensation waveform to obtain sub-band coefficients of each layer.

[0054] Exemplarily, in a 5G NR signal, 4-layer wavelet packet decomposition is performed on the frequency offset compensation waveform to generate 16 sub-band coefficients, and the frequency band numbers are 1 to 16.

[0055] Using a smooth soft threshold function to denoise the sub-band coefficients, the specific formula is: ; where represents the original sub-band coefficient, represents the denoised coefficient, represents the threshold parameter, represents the estimated value of the noise standard deviation.

[0056] The threshold parameter can be dynamically adjusted according to the sub-band number, such as for high-frequency sub-bands ; The noise standard deviation σ is estimated by the median absolute deviation of the sub - band coefficients.

[0057] Exemplarily, for high - frequency sub - bands, such as numbers 13 - 16, λ = 0.3, and the proportion of filtered noise power is ≥80%.

[0058] Retaining the details of the effective signal through a non - linear threshold function can improve the noise suppression rate.

[0059] Pre - define weights according to the importance of the sub - bands in the mobile communication signal. For example, for bands 1 - 4, the weight value can be 0.5; for bands 5 - 8, the weight value is 0.8; for bands 9 - 12, the weight value is 1.0; for bands 13 - 16, the weight value is 0.3.

[0060] Calculate the energy of the denoised sub - band coefficients and weight them. Reference formula: ; where represents energy, represents the weight value of the k - th sub - band, and N represents the number of coefficients.

[0061] Exemplarily, the energy contribution of bands 9 - 12 accounts for 60% of the total, and the proportion of bands 13 - 16 is 5%.

[0062] Differentiated weighting based on the importance of the frequency bands can increase the proportion of the energy contribution of the main carrier frequency band.

[0063] Extract a specified number of sub - bands with the highest energy from all sub - band energies, for example, 8. If the energies of multiple sub - bands are the same, preferentially retain the low - frequency bands.

[0064] Sort the energies of the selected sub - bands according to the band numbers to generate an effective band energy matrix.

[0065] Exemplarily, the energy values of band numbers 7, 8, 9, 10, 11, 12, 5, 6 form the matrix [45, 52, 68, 75, 82, 89, 38, 41].

[0066] Extracting key sub - bands through energy sorting can reduce the subsequent model calculation amount.

[0067] In some embodiments, the protocol - layer signaling anomaly statistics include the IMSI request frequency; adjusting the Doppler frequency offset characteristic weight according to the effective correlation period includes: setting a frequency threshold and a first proportional value; obtaining the sudden increase amount of the IMSI request frequency; when the sudden increase amount of the IMSI request frequency is greater than the frequency threshold, increasing the splitting gain weight of the Doppler frequency offset - related sub - bands to the first proportional value.

[0068] Statistical IMSI request frequency within the effective correlation period and calculate its sudden increase amount compared with the previous period.

[0069] The frequency threshold can be determined according to the minimum frequency of typical IMSI sniffing behavior in pseudo base station attacks, such as being set to 30 times, and the first proportional value can be determined through cross-verification, such as being set to 1.5.

[0070] If the sudden increase in IMSI request frequency is greater than the frequency threshold, the splitting gain weight is increased to 1.5 times the original.

[0071] Only the weights of the sub-bands related to Doppler frequency offset are increased, such as frequency bands numbered 9 - 12, and the weights of other sub-bands remain unchanged, such as the sub-bands related to multipath delay spread.

[0072] Dynamically adjusting the weight based on the sudden increase in IMSI request frequency can improve the sensitivity to instantaneous attacks.

[0073] Combining the frequency threshold and the first proportional value can reduce the misjudgment of low-speed crawling attacks.

[0074] Only adjusting the weights of the sub-bands related to Doppler frequency offset can effectively avoid the degradation of model performance caused by global weight perturbation.

[0075] In some embodiments, the protocol layer signaling anomaly statistics include the number of TA value jumps; adjusting the classification threshold according to the effective correlation period includes: setting the jump threshold and the second proportional value; when the number of TA value jumps is greater than the jump threshold, the splitting gain weight of the sub-bands related to multipath delay spread is attenuated to the second proportional value.

[0076] The number of TA value jumps is statistically counted within the effective correlation period, and a jump is defined as the absolute difference between adjacent TA values exceeding a preset threshold, such as 3 time units.

[0077] Exemplarily, the TA value sequence is [10, 13, 15, 18, 22], the adjacent difference sequence is [3, 2, 3, 4], if the threshold is 3, then the number of jumps is 3 times, and the differences are 3, 3, 4.

[0078] The jump threshold can be determined according to historical statistics, such as being set to 5 times, and the second proportional value can be set to 0.7.

[0079] Dynamically attenuating the weight based on the number of TA value jumps can reduce the impact of multipath interference on the model.

[0080] If the number of TA value jumps is greater than the jump threshold, weight attenuation is performed, and only the weights of the sub-bands related to multipath delay spread are attenuated, such as frequency bands 1 - 4; the weights of other sub-bands remain unchanged.

[0081] Combining the jump threshold and the second proportional value can reduce misjudgment.

[0082] Only adjusting the weights of the sub-bands related to multipath delay spread can effectively avoid the loss of signal features caused by global weight adjustment.

[0083] In some embodiments, the channel environment classification labels at least include Label 1 and Label 2; in the wavelet packet decomposition of the frequency offset compensation waveform, the number of wavelet packet decomposition layers is the first specified number of layers; a second specified number of layers is set, and the second specified number of layers is greater than the first specified number of layers.

[0084] After outputting the channel environment classification label: if the channel environment classification label is Label 1, increase the number of wavelet packet decomposition layers in the next cycle from the first specified number of layers to the second specified number of layers; if the channel environment classification label is Label 2, shorten the sliding window length to the specified window length.

[0085] Label 1 characterizes that the base station signal environment is stable, such as a normal base station, and Label 2 characterizes that the base station signal is abnormal, such as a fake base station.

[0086] The first specified number of layers is the initial number of wavelet packet decomposition layers, such as 4 layers, and the second specified number of layers can be set to 6 layers. The specified window length can be 5 ms. For example, when Label 2 is triggered, the sliding window length is shortened from the default 10 ms to 5 ms.

[0087] If the channel environment classification label output by the fake base station identification module is Label 1, it is determined that the current environment is stable, and the number of wavelet packet decomposition layers in the next cycle is increased from the first specified number of layers to the second specified number of layers.

[0088] The number of decomposition layers is increased to 6 layers, generating 64 sub-band coefficients (originally 16), which can improve the frequency domain resolution.

[0089] If the output label is Label 2, it is determined that there is a risk of a fake base station, and the sliding window length is shortened to the specified window length.

[0090] After the window is shortened, the update frequency of the environmental credibility score is increased from 100 times per second to 200 times, which can improve the real-time performance.

[0091] In some embodiments, the node splitting rules of the random forest classifier in the fake base station identification model include: calculating the protocol anomaly score; setting the anomaly threshold and the mean threshold; if the protocol anomaly score is greater than the anomaly threshold, select the Doppler frequency offset related features for node splitting; if the protocol anomaly score is less than or equal to the anomaly threshold, select the multipath delay spread related features for node splitting; among them, the selected related features satisfy that the historical splitting gain mean under the current protocol anomaly score is greater than the mean threshold.

[0092] The protocol anomaly score S is calculated in the following way: ; where represents the sudden increase in the IMSI request frequency, represents the number of TA value jumps, and α and β represent weight coefficients, with the default α = 0.6 and β = 0.4.

[0093] The abnormal threshold is determined according to historical data statistics. Exceeding this value is regarded as an abnormal high-risk protocol. For example, it is set to 25 points. The mean threshold represents the demarcation value of the historical splitting gain mean, which can be set to 0.2, for example.

[0094] If the protocol abnormal score is greater than the abnormal threshold, select the Doppler frequency offset related features, such as frequency bands 9 - 12, for node splitting; if the protocol abnormal score is less than the abnormal threshold, then select the multipath delay spread related features, such as frequency bands 1 - 4, for node splitting.

[0095] Dynamically switch features through the weighted score of the sudden increase in IMSI request frequency and TA jump to improve the sensitivity of the model to false base station attacks.

[0096] The selected features need to satisfy that the historical splitting gain mean under the current protocol abnormal score is greater than the mean threshold. Exemplarily, if the splitting gain mean of the Doppler frequency offset related features in the scenario where the historical protocol abnormal score is greater than 25 points is 0.25, then splitting is allowed; if the mean is 0.15, then splitting is prohibited.

[0097] Calculate the information gain for the candidate features and select the feature with the maximum gain for splitting.

[0098] Combining the abnormal threshold and the mean threshold for verification can ensure the effectiveness of the splitting features.

[0099] In some embodiments, according to the historical splitting gain mean, filter the Doppler frequency offset related sub - frequency bands and the multipath delay spread related sub - frequency bands. During the training process of each decision tree, calculate the protocol abnormal score in real - time, dynamically switch the splitting features, and determine the optimal parameter combination through grid search.

[0100] In some embodiments, according to the environmental credibility score and the protocol layer signaling anomaly statistics, generate the false base station risk value and the mobile phone terminal geographical coordinates, including: input the environmental credibility score and the protocol layer signaling anomaly statistics into the spatio - temporal gated decision network, output the false base station risk value and the geographical coordinates; perform distance compensation and secondary correction on the geographical coordinates to generate the mobile phone terminal coordinates.

[0101] The spatio - temporal gated decision network includes: Physical branch: Extract the signal stability features from the environmental credibility score through the first fully - connected layer. Set the negative interval slope of the activation function to a specified ratio of the valid association period ratio to generate the physical feature vector.

[0102] Protocol branch: Extract the temporal features from the number of TA jumps and the IMSI anomaly frequency through the gated recurrent unit network to generate the protocol feature vector.

[0103] Fusion layer: Dynamically adjust the weights of the physical branch and the protocol branch according to the proportion of the effective association period, and generate fusion features.

[0104] Second fully connected layer: Output the pseudo base station risk value and the geographical coordinates of the mobile phone end according to the fusion features.

[0105] The first fully connected layer may include 3 fully connected layers, with the number of neurons being 64, 32, and 16 respectively. The activation function can be LeakyReLU, and the negative interval slope is set to 50% of the proportion of the effective association period, outputting a physical feature vector to characterize signal stability.

[0106] The gated recurrent unit (GRU) network structure can be a single-layer GRU, with 32 hidden units. The input is the time series of TA jump times and IMSI anomaly frequencies, the window length is 10ms, the step size is 1ms, and the output is a protocol feature vector to characterize the protocol anomaly time series pattern.

[0107] Calculate the weight of the physical branch according to the proportion of the effective association period, and the weight of the protocol branch is 1 minus the weight of the physical branch.

[0108] The fusion formula can be: ; where represents the fusion feature, represents the weight of the physical branch, represents the physical feature vector, represents the weight of the protocol branch, represents the protocol feature vector.

[0109] The fusion features are input into the second fully connected layer to output the pseudo base station risk value and the geographical coordinates. The pseudo base station risk value can be a scalar between 0 and 1.

[0110] In some embodiments, when training the spatio-temporal gated decision network, the pseudo base station risk value measures the prediction difference using binary cross-entropy loss, and the geographical coordinates measure the coordinate deviation using mean squared error to help the model make accurate predictions.

[0111] The Adam optimizer can be selected, with an initial learning rate of 0.001, a batch size of 128. The learning rate is decreased when the validation loss stagnates, and the L2 regularization coefficient is 0.001 to prevent overfitting. In the forward propagation, the branch weights are dynamically calculated according to the proportion of the effective association period of the sample. For example, when it is 70%, the weight of the physical branch is 0.7 and the weight of the protocol branch is 0.3.

[0112] In some embodiments, distance compensation and secondary correction are performed on geographical coordinates to generate mobile - end coordinates, including: calculating the preliminary distance between the base station and the terminal according to the geographical coordinates; calculating the compensated distance according to the Doppler frequency offset and the preliminary distance; setting a proportion threshold; if the proportion of the effective association period is greater than or equal to the proportion threshold, using the geographical coordinates after the compensated distance as the mobile - end coordinates; if the proportion of the effective association period is less than the proportion threshold, performing secondary correction according to the Doppler frequency offset and the compensated distance, and using the geographical coordinates after the secondary correction as the mobile - end coordinates.

[0113] Obtain the coordinates of known base stations through the operator database, calculate the Euclidean distance between the known base - station coordinates and the mobile end, calculate the compensation amount according to the Doppler frequency offset, and the reference formula is: ; where represents the compensation amount, represents the Euclidean distance between the legal base - station coordinates and the mobile end, c represents the speed of light, represents the Doppler frequency offset, represents the carrier frequency.

[0114] The proportion of the effective association period is the proportion of the effective period in the total signal duration, and the set proportion can be 60%.

[0115] If the proportion of the effective association period is greater than or equal to 60%, directly use the geographical coordinates after the compensated distance as the mobile - end coordinates.

[0116] If the proportion of the effective association period is less than 60%, further adjust the compensated distance according to the Doppler frequency offset, and the reference formula is: , and then, with the base - station coordinates as the center, as the radius, combine multi - base - station triangulation to determine the final mobile - end coordinates.

[0117] The first compensation is based on the Doppler frequency offset, the secondary correction combines the proportion of the effective association period, and the compensation is enhanced in low - quality periods, which can reduce errors. By dynamically switching the correction strategy through the proportion threshold, the balance between accuracy and real - time performance is achieved.

[0118] In some embodiments, as Figure 2 shown, the present disclosure provides a mobile - phone geographical location detection device based on mobile communication signals, including: A radio - frequency front - end module configured to receive the physical - layer radio - frequency signal of the mobile end and generate physical - layer channel impulse response waveform data through down - conversion and analog - to - digital conversion.

[0119] A protocol - parsing chip connected to the radio - frequency front - end module, configured to demodulate protocol - layer signaling data from the physical - layer channel impulse response waveform data and count protocol - layer signaling anomaly indicators.

[0120] The FPGA chip, connected to the radio frequency front-end module, is configured to: dynamically correct and screen the physical layer channel impulse response waveform data to generate a frequency offset compensation waveform; perform wavelet packet decomposition on the frequency offset compensation waveform to generate an effective band energy matrix.

[0121] The main control unit, connected to the FPGA chip and the protocol parsing chip, is configured to: input the effective band energy matrix into the pseudo base station identification model, which is based on a random forest classifier, dynamically adjust the Doppler frequency offset feature weights and classification thresholds according to the effective correlation period, and output the channel environment classification label; perform a sliding window weighted average on the channel environment classification label to generate an environment credibility score; generate a pseudo base station risk value and the mobile phone end geographical coordinates according to the environment credibility score and the protocol layer signaling anomaly statistics.

[0122] It should be noted that in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article or device including the element.

[0123] Although the present disclosure has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A mobile phone geographic location detection system based on mobile communication signals, characterized in that: include: Data receiving module: receives physical layer channel impulse response waveform data and protocol layer signaling data from the mobile phone; Data processing module: dynamically correct and filter the physical layer channel impulse response waveform data to generate a frequency offset compensation waveform; analyze the protocol layer signaling data to generate protocol layer signaling anomaly statistics; Frequency domain feature extraction module: performs wavelet packet decomposition on the frequency offset compensation waveform to generate an effective frequency band energy matrix; Fake base station identification module: The generated effective frequency band energy matrix is ​​input into the fake base station identification model. The fake base station identification model is based on the random forest classifier, and adjusts the Doppler frequency deviation feature weight and classification threshold according to the effective association period, and outputs the channel environment classification label; Comprehensive analysis module: Perform sliding window weighted averaging on channel environment classification labels to generate an environment credibility score; Risk assessment and positioning module: Generates fake base station risk value and mobile phone geographic coordinates based on environmental credibility score and protocol layer signaling anomaly statistics.

2. The mobile phone geographic location detection system based on mobile communication signals according to claim 1 is characterized in that: Dynamically correct and filter the physical layer channel impulse response waveform data to generate a frequency offset compensation waveform, including: Calculate the dynamic parameters of the physical layer channel impulse response waveform to obtain multipath delay spread and Doppler frequency deviation; According to the multipath delay spread, the main path part of the physical layer channel impulse response waveform is dynamically intercepted to generate a delay correction waveform; The waveform is phase compensated according to the Doppler frequency offset to generate a frequency offset compensated waveform.

3. The mobile phone geographic location detection system based on mobile communication signals according to claim 2 is characterized in that: Calculate the time-varying correlation coefficient sequence of the delay correction waveform and the frequency offset compensation waveform; Setting a correlation threshold and a time threshold; screening a time period whose absolute value of the correlation coefficient is greater than the correlation threshold and whose duration is greater than the time threshold in the time-varying correlation coefficient sequence, and generating a valid correlation time period mark; During the effective correlation period, the frequency offset compensation waveform is decomposed by wavelet packets.

4. The mobile phone geographic location detection system based on mobile communication signals according to claim 1, characterized in that: Perform wavelet packet decomposition on the frequency offset compensation waveform to generate an effective frequency band energy matrix, including: Perform wavelet packet decomposition on the frequency offset compensation waveform to obtain the sub-band coefficients of each layer; The smooth soft threshold function is used to denoise the sub-band coefficients, and the sub-band energy is calculated by weight according to the importance of the sub-band in the mobile communication signal; Extract the energy of a specified number of sub-bands to generate an effective sub-band energy matrix.

5. The mobile phone geographic location detection system based on mobile communication signals according to claim 3 is characterized in that: Protocol layer signaling anomaly statistics include IMSI request frequency; The Doppler frequency deviation feature weight is adjusted according to the effective association period, including: Set a frequency threshold and a first ratio value; obtain a sudden increase in the IMSI request frequency; When the IMSI request frequency surge is greater than the frequency threshold, the split gain weight of the Doppler frequency offset related sub-band is increased to a first proportional value.

6. The mobile phone geographic location detection system based on mobile communication signals according to claim 3 is characterized in that: Protocol layer signaling anomaly statistics include the number of TA value jumps; Adjust the classification threshold based on the effective association period, including: Set the jump threshold and the second ratio value; When the number of TA value jumps is greater than the jump threshold, the split gain weight of the multipath delay spread related sub-band is attenuated to a second proportional value.

7. The mobile phone geographic location detection system based on mobile communication signals according to claim 4, characterized in that: The channel environment classification label includes at least label one and label two; in performing wavelet packet decomposition on the frequency offset compensation waveform, the wavelet packet decomposition layer number is a first specified layer number; a second specified layer number is set, and the second specified layer number is greater than the first specified layer number; After outputting the channel environment classification label: if the channel environment classification label is label one, the number of wavelet packet decomposition layers in the next cycle is increased from the first specified number of layers to the second specified number of layers; if the channel environment classification label is label two, the sliding window length is shortened to the specified window length.

8. The mobile phone geographic location detection system based on mobile communication signals according to claim 1, characterized in that: The node splitting rules of the random forest classifier in the fake base station identification model include: Calculate protocol anomaly scores; set anomaly thresholds and mean thresholds; If the protocol anomaly score is greater than the anomaly threshold, the Doppler frequency deviation related features are selected for node splitting; if the protocol anomaly score is less than or equal to the anomaly threshold, the multipath delay spread related features are selected for node splitting; Among them, the selected relevant features satisfy that their historical split gain mean under the current protocol anomaly score is greater than the mean threshold.

9. The mobile phone geographic location detection system based on mobile communication signals according to claim 3, characterized in that: Generate fake base station risk value and mobile phone geographic coordinates based on the environmental credibility score and protocol layer signaling anomaly statistics, including: The environmental credibility score and protocol layer signaling anomaly statistics are input into the spatiotemporal gating decision network, and the risk value and geographic coordinates of the fake base station are output; Perform distance compensation and secondary correction on geographic coordinates to generate mobile phone coordinates; Wherein, the spatiotemporal gating decision network comprises: Physical branch: The signal stability feature is extracted from the environmental credibility score through the first fully connected layer, and the negative interval slope of the activation function is set to the specified proportion of the effective correlation period to generate a physical feature vector; Protocol branch: extract timing features from the TA jump times and IMSI abnormality frequency through the gated recurrent unit network to generate a protocol feature vector; Fusion layer: Dynamically adjust the weights of physical branches and protocol branches according to the proportion of effective association time periods to generate fusion features; The second fully connected layer: outputs the risk value of the fake base station and the geographic coordinates of the mobile phone based on the fusion features.

10. The mobile phone geographic location detection system based on mobile communication signals according to claim 9, characterized in that: Perform distance compensation and secondary correction on geographic coordinates to generate mobile phone coordinates, including: Calculate the preliminary distance between the base station and the terminal according to the geographic coordinates; Calculate the compensated distance according to the Doppler frequency deviation and the preliminary distance; Set a percentage threshold; if the percentage of the effective association period is greater than or equal to the percentage threshold, the coordinates after the distance compensation are used as the coordinates of the mobile phone; if the percentage of the effective association period is less than the percentage threshold, a secondary correction is performed based on the Doppler frequency deviation and the compensated distance, and the geographical coordinates after the secondary correction are used as the coordinates of the mobile phone.

Images