A signal-sensing-based method for quickly detecting a mobile terminal involved in a case

Abstract

This invention relates to the field of baseband communication technology, specifically to a rapid detection method for mobile terminals involved in a case based on signal sensing. The method involves down-converting the captured air interface radio frequency signal into a baseband data stream, extracting the random access preamble sequence and removing the cyclic prefix during the physical layer synchronization phase, performing a Fast Fourier Transform on the preamble sequence to extract the carrier frequency offset estimate and the inter-symbol phase difference sequence, and combining them to generate a baseband physical layer feature vector. This feature vector is then subjected to a time-domain sliding cross-correlation operation with a pre-stored baseband fingerprint database of the involved terminal. When the cross-correlation peak exceeds the decision threshold and the frequency offset change rate is consistent, a detection trigger signal is output and the timestamp is locked. This invention bypasses the higher-layer signaling parsing and encryption cracking stages, directly using baseband waveform features at the physical layer to complete the matching judgment, eliminating the processing delay and decryption failure risk caused by the step-by-step parsing of the protocol stack, and achieving rapid detection of the involved mobile terminal.

Description

Technical Field

[0001] This invention relates to the field of baseband communication technology, and specifically to a rapid detection method for a mobile terminal in question based on signal sensing. Background Technology

[0002] Existing mobile terminal detection schemes typically employ a fake base station deception mechanism. The detection device transmits broadcast control signals with a power exceeding that of a normal base station, forcing mobile terminals within the coverage area to disconnect from the existing network and reselect the detection device. When a terminal initiates a random access request, the detection device receives the air interface radio frequency signal and down-converts it to baseband. It then enters the non-access stratum signaling parsing process to extract the identification information reported by the terminal. Because current communication standards require terminals to encrypt non-access stratum signaling transmission, after obtaining the signaling data, the detection device must call a decryption algorithm library to attempt to crack the encrypted payload, restore the plaintext International Mobile Subscriber Identity (IMSI), and compare this ISI with a local list of suspected mobile terminals. If the comparison matches, the detection is considered successful, and an alarm is output.

[0003] The aforementioned existing technical solutions rely on the complete interaction of higher-level signaling and the successful decryption of the encrypted payload. When the terminal in question uses a high-level security encryption algorithm, the detection device faces a technical barrier to decryption failure at the non-access layer, causing the entire detection process to be directly interrupted and fail. In conventional scenarios where decryption can be successful, the detection system must completely receive and parse the underlying physical layer protocol data units, radio link control layer protocol data units, and non-access layer signaling. The hierarchical parsing and processing of multiple protocol stacks increases the number of data processing levels and the waiting period, causing the detection system to be unable to complete the determination the instant the terminal sends a random access signal, resulting in a response lag problem. Summary of the Invention

[0004] The purpose of this invention is to provide a rapid detection method for mobile terminals involved in a case based on signal perception, which can effectively solve the problems in the background art mentioned above.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A method for rapid detection of involved mobile terminals based on signal sensing, comprising:

[0007] The air interface radio frequency signal captured by the multi-channel radio frequency receiver is down-converted and filtered to convert it into a baseband IQ data stream;

[0008] During the physical layer synchronization phase, the random access preamble sequence of the target frequency band is extracted from the baseband IQ data stream, and the cyclic prefix in the random access preamble sequence is removed.

[0009] The random access preamble sequence after removing the cyclic prefix is ​​processed by fast Fourier transform, and the carrier frequency offset estimate and the inter-symbol phase difference sequence are extracted in the frequency domain. The carrier frequency offset estimate and the inter-symbol phase difference sequence are concatenated and combined in time sequence to generate the baseband physical layer feature vector.

[0010] The baseband physical layer feature vector is subjected to time-domain sliding cross-correlation with the pre-stored baseband fingerprint database of the terminal involved in the case. When the cross-correlation peak exceeds the preset decision threshold and the rate of change of the carrier frequency offset estimate is consistent with the rate of change of frequency offset in the baseband fingerprint database of the terminal involved in the case, a detection trigger signal is output and the timestamp of the corresponding baseband data frame is locked.

[0011] Preferably, the air interface radio frequency signal captured by the multi-channel radio frequency receiver is down-converted and filtered to convert it into a baseband IQ data stream, including:

[0012] The air interface radio frequency signal is converted from analog to digital to obtain a digital intermediate frequency signal;

[0013] The digital intermediate frequency signal is mixed with the local oscillator signal generated by the numerically controlled oscillator to obtain a zero intermediate frequency signal.

[0014] The zero intermediate frequency signal is sequentially subjected to cascaded integral comb filtering and finite-length unit impulse response low-pass filtering to output the baseband IQ data stream;

[0015] During the generation of the baseband IQ data stream, the mean values ​​of the in-phase component and quadrature component of the zero intermediate frequency signal within a preset time period are calculated. The mean values ​​of the in-phase component and quadrature component are used as DC offset. The baseband IQ data stream is then subjected to a subtraction compensation operation using the DC offset. The frequency of the local oscillator signal is dynamically configured according to the center frequency of the target frequency band.

[0016] Preferably, during the physical layer synchronization phase, the random access preamble sequence of the target frequency band is extracted from the baseband IQ data stream, and the cyclic prefix in the random access preamble sequence is removed, including:

[0017] The baseband IQ data stream is cross-correlated with the locally stored physical broadcast channel synchronization sequence to determine the starting position of the physical broadcast channel frame.

[0018] The time domain position of the random access channel in the target frequency band is calculated based on the starting position of the physical broadcast channel frame and the system frame number contained in the physical layer control signaling.

[0019] The random access preamble sequence is extracted from the baseband IQ data stream according to the time domain position;

[0020] Obtain the cyclic prefix length corresponding to the random access preamble sequence, set the data segments in the random access preamble sequence that are within the range of the cyclic prefix length to zero, and output the random access preamble sequence after removing the cyclic prefix.

[0021] Preferably, the random access preamble sequence after removing the cyclic prefix is ​​subjected to Fast Fourier Transform processing to extract the carrier frequency offset estimate and the inter-symbol phase difference sequence in the frequency domain, including:

[0022] Perform a Fast Fourier Transform on the random access preamble sequence after removing the cyclic prefix to obtain a frequency domain sequence;

[0023] Extract the frequency domain response values ​​of two subcarriers with a set index interval from the frequency domain sequence;

[0024] The estimated carrier frequency offset is calculated based on the phase angle of the conjugate multiplication result of the frequency domain response values ​​of the two subcarriers;

[0025] The frequency response values ​​corresponding to adjacent symbols in the frequency domain sequence are multiplied by conjugate, and the phase difference value of the conjugate multiplication result is extracted. The phase difference values ​​corresponding to all adjacent symbols in the random access preamble sequence are arranged in time order to generate the inter-symbol phase difference sequence.

[0026] Preferably, the baseband physical layer feature vector is generated by sequentially concatenating the carrier frequency offset estimate and the inter-symbol phase difference sequence, including:

[0027] The estimated carrier frequency offset is normalized and quantized to generate a frequency offset quantization sequence;

[0028] The inter-symbol phase difference sequence is subjected to symbol bit value mapping processing to generate a phase difference quantization sequence;

[0029] The frequency offset quantization sequence and the phase difference quantization sequence are alternately interpolated according to the timestamp order of the random access preamble sequence to generate an intermediate splicing sequence;

[0030] The intermediate spliced ​​sequence is windowed and truncated to extract the portion of the intermediate spliced ​​sequence that corresponds to the valid data segment of the random access preamble sequence. The truncated sequence is then used as the baseband physical layer feature vector.

[0031] Preferably, performing a time-domain sliding cross-correlation operation between the baseband physical layer feature vector and a pre-stored baseband fingerprint database of the terminal involved in the case includes:

[0032] Multiple reference feature vectors of the terminals involved in the case were retrieved from the baseband fingerprint database of the terminals involved in the case;

[0033] The baseband physical layer feature vector is moved by a preset sliding step size, and the sum of the dot products of the moved baseband physical layer feature vector and each of the terminal reference feature vectors involved in the case is calculated to obtain multiple cross-correlation operation results.

[0034] The maximum value among the multiple cross-correlation results is extracted as the cross-correlation peak value;

[0035] If the cross-correlation peak value exceeds the preset decision threshold, the frequency offset change rate corresponding to the reference feature vector of the terminal involved in the case that generated the cross-correlation peak value is obtained, and it is determined whether the difference between the change rate of the carrier frequency offset estimate and the obtained frequency offset change rate is within a preset error range. If it is within the preset error range, the detection trigger signal is output.

[0036] Preferably, the frequency of the local oscillator signal is dynamically configured according to the center frequency of the target frequency band, including:

[0037] Scan the frequency band range corresponding to the target frequency band to obtain the signal power spectral density at each frequency point within the frequency band range;

[0038] The frequency points in the signal power spectral density that are greater than a preset power threshold are identified as the main interference frequency points;

[0039] Calculate the frequency difference between the center frequency of the target frequency band and the main interference frequency point, and adjust the control word of the numerically controlled oscillator according to the frequency difference so that the frequency of the adjusted local oscillator signal deviates from the main interference frequency point;

[0040] Using the baseband IQ data stream after the subtraction compensation operation, cross-correlation-based time delay estimation is performed on the baseband IQ data streams corresponding to different channels in the multi-channel RF receiver, and time alignment operation is performed on the baseband IQ data streams of different channels according to the time delay estimation results.

[0041] Preferably, after outputting the random access preamble sequence after removing the cyclic prefix, the method further includes:

[0042] Extract the reference signal configuration parameters contained in the physical layer control signaling;

[0043] The reference signal sequence is separated from the random access preamble sequence after removing the cyclic prefix according to the reference signal configuration parameters;

[0044] Channel impulse response estimation is performed using the reference signal sequence to obtain the channel impulse response sequence;

[0045] The number of multipath components and the arrival time of each multipath component are determined based on the power delay distribution of the channel impulse response sequence.

[0046] When the number of multipath components exceeds a preset threshold, frequency domain equalization filter coefficients are constructed based on the channel impulse response sequence;

[0047] The frequency domain equalization filter coefficients are used to perform a frequency domain linear equalization operation on the random access preamble sequence after removing the cyclic prefix.

[0048] Preferably, after calculating the carrier frequency offset estimate based on the phase angle of the conjugate multiplication result of the frequency domain response values ​​of the two subcarriers, the method further includes:

[0049] The estimated carrier frequency offset is processed by separating the fractional and integer multiples of frequency offset.

[0050] Phase rotation compensation is performed on the frequency domain sequence using the separated fractional frequency offset;

[0051] The frequency domain sequence after phase rotation compensation is subjected to inverse fast Fourier transform to obtain the time domain compensated sequence.

[0052] Autocorrelation is performed on the first and last data segments of the time-domain compensation sequence to calculate the integer multiples of the frequency offset;

[0053] The integer multiple frequency offset and the fractional multiple frequency offset are superimposed to generate a refined carrier frequency offset;

[0054] The inter-symbol phase difference sequence is recalculated using the refined carrier frequency offset, and the phase jump points caused by noise in the recalculated inter-symbol phase difference sequence are smoothed by linear interpolation.

[0055] Preferably, after outputting the detection trigger signal and locking the timestamp of the corresponding baseband data frame, the method further includes:

[0056] Based on the locked timestamp, extract consecutive data frames from the baseband IQ data stream that are within a preset delay range after the timestamp;

[0057] The continuous data frames are demodulated and channel decoded to extract Media Access Control (MAC) protocol data units.

[0058] Parse the header field of the Media Access Control Layer Protocol data unit and extract the wireless network temporary identifier from the header field;

[0059] The temporary wireless network identifier is hash-compared with a pre-stored database of identifiers for terminals involved in the case.

[0060] If the hash comparison results are consistent, a final confirmation instruction is generated, and the final confirmation instruction is bound and stored with the locked timestamp.

[0061] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0062] 1. This invention downconverts the air interface radio frequency signal into an in-phase quadrature baseband data stream. During the physical layer synchronization phase, it directly extracts the random access preamble sequence and removes the cyclic prefix. A fast Fourier transform is performed on the preamble sequence to extract the carrier frequency offset estimate and the inter-symbol phase difference sequence. These are combined to generate a baseband physical layer feature vector and subjected to time-domain sliding cross-correlation with a pre-stored fingerprint database. When the cross-correlation peak value and the frequency offset change rate are consistent, a probe signal is output and the timestamp is locked. This invention bypasses the non-access layer signaling interaction and encryption parsing stages, directly utilizing the hardware characteristics of the physical layer baseband waveform for matching and determination. This eliminates the processing delay and failure risk caused by higher-layer decryption and shortens the detection cycle.

[0063] 2. By calculating the mean values ​​of the in-phase and quadrature components of the zero-IF signal for DC offset compensation, and dynamically adjusting the local oscillator frequency to deviate from the interference source based on the main interference frequency, and combining multi-channel cross-correlation delay estimation for time alignment, the signal quality and multi-channel consistency of the in-phase and quadrature baseband data stream are improved. By separating the reference signal sequence for channel impulse response estimation and constructing frequency domain equalization filter coefficients to perform linear equalization, inter-symbol interference caused by multipath channels is overcome. By separating the carrier frequency offset estimate for fractional and integer multiple compensation, and performing linear interpolation smoothing on the noise phase jump points in the phase difference sequence, the extraction errors of frequency offset and phase features are reduced. After outputting the detection signal, by extracting the temporary wireless network identifier in subsequent continuous data frames and performing hash comparison with the identifier database of the terminal involved, a secondary verification from physical layer coarse screening to media access control layer precise authentication is achieved, ensuring the determinism of the final detection result. Attached Figure Description

[0064] Figure 1 This is a flowchart illustrating the overall process of rapid detection of the mobile terminal involved in a case based on signal perception according to the present invention.

[0065] Figure 2 This is a flowchart of the down-conversion and DC offset compensation process of the present invention;

[0066] Figure 3 This is a flowchart of the physical layer synchronization and random access preamble processing of the present invention;

[0067] Figure 4 This is a flowchart of the channel estimation and frequency domain equalization process of the present invention;

[0068] Figure 5 This is a flowchart of the carrier frequency offset refinement estimation and baseband physical layer feature vector construction of the present invention;

[0069] Figure 6This is a flowchart of the baseband fingerprint database matching and MAC layer secondary verification process of the present invention. Detailed Implementation

[0070] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0071] Please refer to Figure 1 This embodiment provides a rapid detection method for a mobile terminal involved in a case based on signal sensing. A multi-channel RF receiver operates in continuous reception mode within the target communication frequency band. The receiver's RF front-end is configured with a bandpass filter network that matches the frequency range of the target frequency band to capture air interface RF signals propagating in space within the coverage area. Specifically, the multi-channel RF receiver transmits the captured air interface RF signals to a baseband processing unit. The baseband processing unit's built-in analog-to-digital conversion module, digital down-conversion module, and filtering module work together to perform down-conversion and filtering on the air interface RF signals, converting the analog signals of the RF band into a baseband in-phase quadrature data stream, i.e., a baseband IQ data stream.

[0072] Furthermore, during the physical layer synchronization phase, the synchronization detection module built into the baseband processing unit performs frame synchronization and symbol synchronization processing on the generated baseband IQ data stream, extracts the random access preamble sequence corresponding to the target frequency band from the baseband IQ data stream, and simultaneously performs cyclic prefix removal operation on the extracted random access preamble sequence to eliminate the impact of inter-symbol interference caused by the cyclic prefix on subsequent frequency domain processing.

[0073] In this embodiment, the frequency domain transformation module built into the baseband processing unit performs fast Fourier transform processing on the random access preamble sequence after cyclic prefix removal, mapping the time-domain preamble sequence to the frequency domain. The frequency domain transformation module extracts the carrier frequency offset estimate and generates the inter-symbol phase difference sequence in the frequency domain. The feature vector construction module concatenates and combines the carrier frequency offset estimate and the inter-symbol phase difference sequence in time sequence to generate the baseband physical layer feature vector.

[0074] Specifically, the computational process of Fast Fourier Transform (FFT) satisfies the following expression:

[0075] ;

[0076] in, This represents the nth time-domain sampling point of the random access preamble sequence after removing the cyclic prefix, where N is the number of points in the Fast Fourier Transform. The value of N is consistent with the effective data segment length of the random access preamble sequence. This represents the frequency domain sequence value corresponding to the k-th subcarrier.

[0077] The frequency domain transformation module extracts the frequency domain response values ​​of two subcarriers with a set index interval from the frequency domain sequence. It then calculates the carrier frequency offset estimate based on the phase angle of the conjugate multiplication result of the two subcarrier frequency domain response values. The calculation process for the carrier frequency offset estimate satisfies the following expression:

[0078] ;

[0079] in, The subcarrier spacing of the target communication system. The index interval between two subcarriers. Let k+m be the frequency domain response value of the subcarrier. The conjugate of the frequency domain response value of the subcarrier at index k. For phase angle taking operation, This is the calculated carrier frequency offset estimate.

[0080] The frequency domain transformation module performs conjugate multiplication on the frequency domain response values ​​corresponding to adjacent symbols in the frequency domain sequence, extracts the phase difference value of the conjugate multiplication result, and arranges the phase difference values ​​corresponding to all adjacent symbols in the randomly accessed preamble sequence in time order to generate an inter-symbol phase difference sequence. The calculation process of the phase difference value satisfies the following expression:

[0081] ;

[0082] in, Let be the frequency domain response value of the k-th subcarrier corresponding to the t-th symbol. Let be the frequency domain response value of the k-th subcarrier corresponding to the (t+1)-th symbol. for The conjugate of , where T is the total number of symbols contained in the random access preamble sequence, Let t be the phase difference value corresponding to the t-th time position. All phase differences are arranged in time sequence to form an inter-symbol phase difference sequence.

[0083] The feature vector construction module performs normalization and quantization processing on the carrier frequency offset estimate to generate a frequency offset quantization sequence. The normalization and quantization process satisfies the following expression:

[0084] ;

[0085] in, Let be the estimated carrier frequency offset at time t. This is the maximum value of the carrier frequency offset estimate within the preset observation window. Q represents the minimum estimated carrier frequency offset within the preset observation window, where Q is the number of quantization bits. This is a rounding operation. Let be the frequency offset quantization value at time t. All frequency offset quantization values ​​are arranged in time sequence to form a frequency offset quantization sequence.

[0086] Furthermore, the feature vector construction module performs symbol bit value mapping processing on the inter-symbol phase difference sequence to generate a phase difference quantization sequence. The frequency offset quantization sequence and the phase difference quantization sequence are interpolated alternately according to the timestamp order of the random access preamble sequence to generate an intermediate splicing sequence. The intermediate splicing sequence is windowed and truncated, and the part of the intermediate splicing sequence corresponding to the effective data segment of the random access preamble sequence is extracted. The truncated sequence is used as the baseband physical layer feature vector.

[0087] The feature matching module performs a time-domain sliding cross-correlation operation between the generated baseband physical layer feature vector and the pre-stored baseband fingerprint database of the involved terminals. The time-domain sliding cross-correlation operation process satisfies the following expression:

[0088] ;

[0089] in, Let L be the i-th element of the baseband physical layer feature vector, and L be the length of the baseband physical layer feature vector. The sliding offset is The first reference feature vector of the terminal involved in the case There are elements, where M is the length of the reference feature vector of the terminal involved in the case. Offset The corresponding cross-correlation results.

[0090] Specifically, the feature matching module extracts the maximum value among multiple cross-correlation calculation results as the cross-correlation peak value. If the cross-correlation peak value exceeds the preset decision threshold, the frequency offset change rate corresponding to the reference feature vector of the terminal involved in the case that generates the cross-correlation peak value is obtained. The difference between the change rate of the carrier frequency offset estimate and the obtained frequency offset change rate is judged to be within the preset error range. If it is within the preset error range, the detection trigger signal is output, and the timestamp of the corresponding baseband data frame is locked.

[0091] Table 1. Parameter composition and timing correspondence of the baseband physical layer eigenvectors

[0092]

[0093] In this embodiment, Table 1 clarifies the temporal position, data length, and quantization configuration of each intermediate parameter during the generation of the baseband physical layer feature vector, unifies the dimension and format of the feature vector, ensures the temporal consistency and dimensional matching between the feature vector and the reference feature vector in the baseband fingerprint database of the terminal involved in the case, and provides standardized data input for subsequent temporal sliding cross-correlation operations.

[0094] In this embodiment, rapid physical layer detection of the mobile terminal in question is achieved through down-conversion and filtering of the air interface radio frequency signal, extraction of the random access preamble sequence and cyclic prefix removal during the physical layer synchronization stage, extraction of carrier frequency offset and phase difference features after frequency domain transformation, construction of baseband physical layer feature vectors, and time-domain sliding cross-correlation matching and judgment with the baseband fingerprint database of the terminal in question. The entire processing flow does not require complete interaction of higher-layer signaling and parsing of encrypted payloads, and directly completes the detection and judgment based on the baseband hardware characteristics of the physical layer, eliminating the processing delay and process failure risk caused by the higher-layer decryption link, and shortening the entire process cycle of detection and processing.

[0095] In a preferred embodiment, reference Figure 2 When the baseband processing unit performs down-conversion and filtering on the air interface RF signal captured by the multi-channel RF receiver, it first performs analog-to-digital conversion on the air interface RF signal to obtain a digital intermediate frequency (IF) signal. Specifically, the baseband processing unit mixes the digital IF signal with the local oscillator signal generated by the numerically controlled oscillator to obtain a zero IF signal. The mixing process satisfies the following expression:

[0096] ;

[0097] in, This is the digital intermediate frequency signal at the nth sampling point after analog-to-digital conversion. The frequency of the local oscillator signal generated by the numerically controlled oscillator. The sampling period for analog-to-digital conversion. The in-phase component of the zero intermediate frequency signal. These are the orthogonal components of the zero intermediate frequency signal.

[0098] Furthermore, the baseband processing unit sequentially performs cascaded integral comb filtering and finite-length unit impulse response low-pass filtering on the zero intermediate frequency signal, and outputs the baseband IQ data stream after the filtering process is completed.

[0099] In this embodiment, during the generation of the baseband IQ data stream, the baseband processing unit calculates the mean value of the in-phase component and the mean value of the quadrature component of the zero intermediate frequency signal within a preset time period. The mean values ​​of the in-phase component and the quadrature component are used as the DC offset. The DC offset is then used to perform a subtraction compensation operation on the baseband IQ data stream. The DC offset compensation calculation process satisfies the following expression:

[0100] ;

[0101] in, This is the average value of the in-phase components within a preset time period. The mean of the orthogonal components within a preset time period. This is the in-phase baseband data after DC compensation. This is the baseband data of the quadrature components after DC compensation. The length of the preset time period is configured according to the stability of the signal.

[0102] Specifically, the frequency of the local oscillator signal generated by the CNC oscillator is dynamically configured according to the center frequency of the target frequency band. The baseband processing unit first scans the frequency band range corresponding to the target frequency band, obtains the signal power spectral density of each frequency point within the frequency band range, and determines the frequency points with signal power spectral density greater than the preset power threshold as the main interference frequency points. The frequency distance between the center frequency of the target frequency band and the main interference frequency points is calculated, and the control word of the CNC oscillator is adjusted according to the frequency distance so that the frequency of the adjusted local oscillator signal deviates from the main interference frequency points, thereby avoiding the influence of the main interference frequency points on the down-conversion processing.

[0103] Furthermore, the baseband processing unit utilizes the baseband IQ data stream after the subtraction compensation operation to perform cross-correlation-based time delay estimation on the baseband IQ data streams corresponding to different channels in the multi-channel RF receiver. Based on the time delay estimation results, it performs time alignment operations on the baseband IQ data streams of different channels. The time delay estimation process is implemented based on cross-correlation peak detection. The baseband processing unit uses the baseband IQ data stream of one channel as a reference sequence, calculates the cross-correlation results between the baseband IQ data streams of other channels and the reference sequence, extracts the offset corresponding to the cross-correlation peak as the time delay difference between different channels, and performs shift compensation on the baseband IQ data stream of the corresponding channel based on the time delay difference to complete the time alignment of the multi-channel data.

[0104] Table 2. Local Oscillator Signal Frequency Configuration and Interference Avoidance Parameters for Different Target Frequency Bands

[0105]

[0106] In this embodiment, Table 2 provides the dynamic configuration parameters of the local oscillator signal frequency after the main interference frequency point is detected under different target frequency bands, clarifies the correspondence between the CNC oscillator control word and the local oscillator frequency, and gives the frequency offset adjustment range under different configurations, verifying the avoidance effect of dynamic adjustment of local oscillator frequency on the main interference frequency point, and ensuring the signal-to-noise ratio of the baseband IQ data stream after down-conversion processing.

[0107] In this embodiment, the complete conversion process from radio frequency signal to baseband IQ data stream is completed through analog-to-digital conversion, digital mixing, cascaded filtering, and low-pass filtering of the air interface radio frequency signal. Simultaneously, DC component interference caused by local oscillator leakage and mixing nonlinearity under the zero-IF architecture is eliminated through DC offset calculation and subtraction compensation. By scanning the target frequency band and detecting the main interference frequency, the control word of the numerically controlled oscillator is dynamically adjusted to change the local oscillator signal frequency, effectively avoiding the main interference frequency within the band. Combined with cross-correlation delay estimation and time alignment operations of multi-channel baseband data, the signal quality of the baseband IQ data stream and the consistency of multi-channel data are improved, providing a stable and high-quality data input foundation for subsequent physical layer synchronization detection and feature extraction processing.

[0108] In another preferred embodiment, reference Figure 3 When the baseband processing unit extracts the random access preamble sequence of the target frequency band from the baseband IQ data stream during the physical layer synchronization phase, it first performs a cross-correlation operation between the baseband IQ data stream and the locally stored physical broadcast channel synchronization sequence to determine the start position of the physical broadcast channel frame. Specifically, the physical broadcast channel synchronization sequence consists of the primary synchronization sequence and the secondary synchronization sequence specified in the target communication standard. The baseband processing unit performs a sliding cross-correlation operation between the baseband IQ data stream and the locally stored primary synchronization sequence, extracts the time-domain position corresponding to the cross-correlation peak as the primary synchronization point, completes the detection of the secondary synchronization sequence based on the primary synchronization point, and determines the physical cell identifier and the start position of the physical broadcast channel frame based on the detection result of the secondary synchronization sequence.

[0109] Furthermore, the baseband processing unit calculates the time-domain position of the random access channel in the target frequency band based on the start position of the physical broadcast channel frame and the system frame number contained in the physical layer control signaling. The physical layer control signaling is transmitted through the physical downlink control channel. The baseband processing unit demodulates and decodes the signal of the physical downlink control channel, extracts the system frame number and the configuration parameters of the random access channel. The configuration parameters of the random access channel include the time-domain period, the time-domain start symbol position, and the frequency-domain resource position of the random access channel. Combining the start position of the physical broadcast channel frame and the system frame number, the accurate start and end positions of the random access channel in the time domain are calculated.

[0110] In this embodiment, the baseband processing unit extracts the random access preamble sequence from the baseband IQ data stream according to the calculated time-domain position, obtains the cyclic prefix length corresponding to the random access preamble sequence, and the value of the cyclic prefix length is determined by the random access preamble format specified in the target communication standard. The baseband processing unit sets the data segments in the random access preamble sequence that are within the range of the cyclic prefix length to zero, and outputs the random access preamble sequence after removing the cyclic prefix.

[0111] Table 3. Parameters for truncating and removing the cyclic prefix of the random access preamble sequence under different frame structure configurations.

[0112]

[0113] In this embodiment, Table 3 clarifies the cyclic prefix length, total preamble sequence length, and effective data segment truncation parameters corresponding to different random access preamble formats in the target communication standard. This provides a quantitative boundary basis for the cyclic prefix removal operation, ensures the accuracy of effective data segment truncation, and avoids the impact of inter-symbol interference caused by cyclic prefix residue on subsequent frequency domain transformation processing.

[0114] Further, refer to Figure 4 After outputting the random access preamble sequence with the cyclic prefix removed, the baseband processing unit extracts the reference signal configuration parameters contained in the physical layer control signaling. These parameters include the time-domain position, frequency-domain position, and sequence generation parameters of the reference signal. Based on these parameters, the reference signal sequence is separated from the cyclic preamble sequence. Channel impulse response estimation is then performed using this reference signal sequence to obtain the channel impulse response sequence. The channel impulse response estimation process satisfies the following expression:

[0115] ;

[0116] in, The frequency domain values ​​of the received reference signal sequence, The standard frequency domain values ​​of the locally stored reference signal sequence. For the inverse fast Fourier transform operation, N is the number of points in the inverse fast Fourier transform. This is the channel impulse response value corresponding to the nth tap.

[0117] Specifically, the baseband processing unit determines the number of multipath components and the arrival time of each multipath component based on the power delay distribution of the channel impulse response sequence. The power delay distribution is calculated by squaring the magnitude of the channel impulse response sequence. The baseband processing unit performs peak detection on the power delay distribution, identifying peaks whose power exceeds a preset power threshold as valid multipath components, counting the number of valid multipath components, and recording the time-domain position of each valid multipath component as its arrival time. When the number of multipath components exceeds a preset threshold, the baseband processing unit constructs frequency-domain equalization filter coefficients based on the channel impulse response sequence. The calculation process of the frequency-domain equalization filter coefficients satisfies the following expression:

[0118] ;

[0119] in, The frequency domain response value of the channel impulse response sequence. for conjugate, This is a noise power estimate, calculated by taking the average power of the noise region in the channel impulse response sequence. These are the frequency domain equalization filter coefficients corresponding to the k-th subcarrier.

[0120] Furthermore, the baseband processing unit performs a frequency domain linear equalization operation on the random access preamble sequence after removing the cyclic prefix using the frequency domain equalization filter coefficients. Specifically, it first performs a fast Fourier transform on the random access preamble sequence to obtain a frequency domain sequence, multiplies the frequency domain sequence with the frequency domain equalization filter coefficients to obtain an equalized frequency domain sequence, and performs an inverse fast Fourier transform on the equalized frequency domain sequence to obtain a time-domain equalized random access preamble sequence.

[0121] In this embodiment, the precise location of the physical broadcast channel frame start position is achieved through cross-correlation detection of the physical broadcast channel synchronization sequence. Combined with the system frame number and random access channel configuration parameters in the physical layer control signaling, the accurate calculation of the random access channel time-domain position and the precise extraction of the random access preamble sequence are completed. Simultaneously, the cyclic prefix is ​​eliminated based on the cyclic prefix length corresponding to the random access preamble format. Through the separation of the reference signal sequence and channel impulse response estimation, the multipath component distribution characteristics of the wireless channel are obtained. When the number of multipath components exceeds a preset threshold, frequency domain equalization filter coefficients under the minimum mean square error criterion are constructed. Frequency domain linear equalization is performed on the random access preamble sequence, eliminating inter-symbol interference caused by the multipath channel, improving the signal-to-noise ratio of the random access preamble sequence, and ensuring the accuracy of subsequent frequency domain feature extraction.

[0122] In a preferred embodiment, reference Figure 5 The baseband processing unit performs Fast Fourier Transform (FFT) processing on the random access preamble sequence after removing the cyclic prefix. When extracting the carrier frequency offset estimate and the inter-symbol phase difference sequence in the frequency domain, the FFT is first performed on the random access preamble sequence after removing the cyclic prefix to obtain the frequency domain sequence. The frequency domain response values ​​of two subcarriers with a set index interval are extracted from the frequency domain sequence. The carrier frequency offset estimate is calculated based on the phase angle of the conjugate multiplication result of the frequency domain response values ​​of the two subcarriers. The frequency domain response values ​​corresponding to adjacent symbols in the frequency domain sequence are conjugate multiplied, and the phase difference value of the conjugate multiplication result is extracted. The phase difference values ​​corresponding to all adjacent symbols in the random access preamble sequence are arranged in time sequence to generate the inter-symbol phase difference sequence.

[0123] Furthermore, after calculating the carrier frequency offset estimate based on the phase angle of the conjugate multiplication result of the frequency domain response values ​​of the two subcarriers, the baseband processing unit performs fractional and integer frequency offset separation processing on the carrier frequency offset estimate. The separated fractional frequency offset is then used to perform phase rotation compensation on the frequency domain sequence. The phase rotation compensation operation process satisfies the following expression:

[0124] ;

[0125] in, The fractional octave offset value is the value that has been separated. The original frequency domain sequence, This is the frequency domain sequence after phase rotation compensation. The sampling period is k, and the subcarrier index is k.

[0126] Specifically, the baseband processing unit performs an inverse fast Fourier transform on the frequency domain sequence after phase rotation compensation to obtain a time-domain compensated sequence. It then performs autocorrelation calculations on the first and last data segments of the time-domain compensated sequence to calculate integer multiples of the frequency offset. The calculation process for integer multiples of the frequency offset satisfies the following expression:

[0127] ;

[0128] in, The length of the cyclic prefix. This represents the nth sampling point of the time-domain compensated sequence after phase rotation compensation, where N is the number of points in the Fast Fourier Transform. For subcarrier spacing, It is an integer multiple of the frequency offset. This refers to the operation that retrieves the index corresponding to the maximum value.

[0129] Furthermore, the baseband processing unit superimposes the integer multiples and fractional multiples of the frequency offset to generate a refined carrier frequency offset. It then recalculates the inter-symbol phase difference sequence using this refined carrier frequency offset and performs linear interpolation smoothing on the noise-induced phase jump points in the recalculated inter-symbol phase difference sequence. Specifically, the baseband processing unit performs a first-order difference operation on the inter-symbol phase difference sequence, extracts the positions where the absolute value of the first-order difference result exceeds a preset jump threshold as phase jump points, and uses the two adjacent effective phase values ​​before and after the phase jump point as a reference to perform linear interpolation replacement on the values ​​at the phase jump point, thus completing the smoothing of the phase difference sequence.

[0130] In this embodiment, when the baseband processing unit concatenates the carrier frequency offset estimate and the inter-symbol phase difference sequence in a time sequence to generate the baseband physical layer feature vector, it performs normalization and quantization processing on the carrier frequency offset estimate to generate a frequency offset quantization sequence, performs symbol bit value mapping processing on the inter-symbol phase difference sequence to generate a phase difference quantization sequence, and performs alternating interpolation processing on the frequency offset quantization sequence and the phase difference quantization sequence according to the timestamp order of the random access preamble sequence to generate an intermediate concatenated sequence. The intermediate concatenated sequence is then windowed and truncated, and the portion of the intermediate concatenated sequence corresponding to the effective data segment of the random access preamble sequence is extracted. This truncated sequence is used as the baseband physical layer feature vector. Specifically, the windowing process uses a Hamming window function, and the length of the window function is consistent with the length of the effective data segment of the random access preamble sequence to reduce the impact of spectrum leakage on the feature vector.

[0131] Furthermore, when the baseband processing unit performs time-domain sliding cross-correlation calculations on the baseband physical layer feature vectors and the pre-stored baseband fingerprint database of the terminals involved in the case, it retrieves multiple reference feature vectors of the terminals involved in the case from the baseband fingerprint database, moves the baseband physical layer feature vectors with a preset sliding step size, calculates the sum of the dot products of the moved baseband physical layer feature vectors and each reference feature vector of the terminals involved in the case, and obtains multiple cross-correlation calculation results. The maximum value among the multiple cross-correlation calculation results is extracted as the cross-correlation peak value. If the cross-correlation peak value exceeds a preset decision threshold, the frequency offset change rate corresponding to the reference feature vector of the terminal involved in the case that generated the cross-correlation peak value is obtained. It is then determined whether the difference between the change rate of the estimated carrier frequency offset and the obtained frequency offset change rate is within a preset error range. If it is within the preset error range, a detection trigger signal is output. The value of the preset decision threshold is configured based on the energy of the reference feature vector of the terminals involved in the case through normalization, and the value of the preset error range is configured based on the carrier frequency offset stability of the target communication system.

[0132] Table 4 Reference Feature Vector Configuration Parameters for the Baseband Fingerprint Database of the Terminals Involved in the Case

[0133]

[0134] In this embodiment, Table 4 provides the reference feature vector configuration parameters corresponding to different terminals involved in the case in the baseband fingerprint database of the terminals involved in the case. It clarifies the length of the reference feature vector, the benchmark value of the frequency offset change rate, the preset decision threshold of the cross-correlation decision and the preset error range of the frequency offset change rate consistency verification. This provides a unified benchmark for the quantitative decision in the feature matching process and ensures the accuracy of the detection decision and a low false alarm rate.

[0135] Further, refer to Figure 6After outputting a probe trigger signal and locking the timestamp of the corresponding baseband data frame, the baseband processing unit uses the locked timestamp as a reference to extract continuous data frames within a preset delay range following the timestamp from the baseband IQ data stream. The baseband processing unit demodulates and decodes the continuous data frames, extracts Media Access Control (MAC) protocol data units, parses the header field of the MAC protocol data units, and extracts the Radio Network Temporary Identifier (RANI) from the header field.

[0136] Specifically, the baseband processing unit performs a hash comparison between the extracted temporary wireless network identifier and the pre-stored identifier database of the terminal involved in the case. The hash comparison operation satisfies the following expression:

[0137] ;

[0138] in, For the extracted temporary identifier of the wireless network, For locked timestamps, This is a byte concatenation operation. This is a SHA-256 hash operation. The generated hash comparison value is then compared with the hash value pre-stored in the identification database of the terminal involved in the case. If the hash comparison results match, a final confirmation instruction is generated. The final confirmation instruction is bound to the locked timestamp and stored. At the same time, the original data of the corresponding baseband data frame is recorded to provide data support for subsequent tracing.

[0139] In this embodiment, by separating the fractional and integer multiples of the estimated carrier frequency offset, a refined estimation and phase rotation compensation of the carrier frequency offset are achieved, reducing the error in frequency offset estimation. Simultaneously, linear interpolation smoothing is performed on the noise phase jump points in the inter-symbol phase difference sequence, eliminating the impact of noise-induced phase jumps and improving the discriminability and stability of the baseband physical layer feature vector. A rapid matching of the baseband physical layer feature vector and the reference feature vector of the terminal in question is achieved through time-domain sliding cross-correlation. Combined with cross-correlation peak threshold judgment and frequency offset change rate consistency verification, a rapid coarse screening of the physical layer is realized, reducing the probability of false alarms. After outputting the detection trigger signal, the temporary identifier of the wireless network is extracted through demodulation and decoding of subsequent continuous data frames and parsing of Media Access Control layer protocol data units. A second precise confirmation is achieved by combining this with a hash comparison of the pre-stored terminal identifier library. Simultaneously, the final confirmation command is bound and stored with the locked timestamp, ensuring the accuracy, determinism, and traceability of the detection results.

Claims

1. A method for rapid detection of a mobile terminal involved in a case based on signal sensing, characterized in that, include: The air interface radio frequency signal captured by the multi-channel radio frequency receiver is down-converted and filtered to convert it into a baseband IQ data stream; During the physical layer synchronization phase, the random access preamble sequence of the target frequency band is extracted from the baseband IQ data stream, and the cyclic prefix in the random access preamble sequence is removed. The random access preamble sequence after removing the cyclic prefix is ​​processed by fast Fourier transform, and the carrier frequency offset estimate and the inter-symbol phase difference sequence are extracted in the frequency domain. The carrier frequency offset estimate and the inter-symbol phase difference sequence are concatenated and combined in time sequence to generate the baseband physical layer feature vector. The baseband physical layer feature vector is subjected to time-domain sliding cross-correlation with the pre-stored baseband fingerprint database of the terminal involved in the case. When the cross-correlation peak exceeds the preset decision threshold and the rate of change of the carrier frequency offset estimate is consistent with the rate of change of frequency offset in the baseband fingerprint database of the terminal involved in the case, a detection trigger signal is output and the timestamp of the corresponding baseband data frame is locked.

2. The method for rapid detection of a mobile terminal involved in a case based on signal perception according to claim 1, characterized in that, The air interface RF signal captured by the multi-channel RF receiver is down-converted and filtered to convert it into a baseband IQ data stream, including: The air interface radio frequency signal is converted from analog to digital to obtain a digital intermediate frequency signal; The digital intermediate frequency signal is mixed with the local oscillator signal generated by the numerically controlled oscillator to obtain a zero intermediate frequency signal. The zero intermediate frequency signal is sequentially subjected to cascaded integral comb filtering and finite-length unit impulse response low-pass filtering to output the baseband IQ data stream; During the generation of the baseband IQ data stream, the mean values ​​of the in-phase component and quadrature component of the zero intermediate frequency signal within a preset time period are calculated. The mean values ​​of the in-phase component and quadrature component are used as DC offset. The baseband IQ data stream is then subjected to a subtraction compensation operation using the DC offset. The frequency of the local oscillator signal is dynamically configured according to the center frequency of the target frequency band.

3. The method for rapid detection of a mobile terminal involved in a case based on signal perception according to claim 1, characterized in that, During the physical layer synchronization phase, the random access preamble sequence of the target frequency band is extracted from the baseband IQ data stream, and the cyclic prefix in the random access preamble sequence is removed, including: The baseband IQ data stream is cross-correlated with the locally stored physical broadcast channel synchronization sequence to determine the starting position of the physical broadcast channel frame. The time domain position of the random access channel in the target frequency band is calculated based on the starting position of the physical broadcast channel frame and the system frame number contained in the physical layer control signaling. The random access preamble sequence is extracted from the baseband IQ data stream according to the time domain position; Obtain the cyclic prefix length corresponding to the random access preamble sequence, set the data segments in the random access preamble sequence that are within the range of the cyclic prefix length to zero, and output the random access preamble sequence after removing the cyclic prefix.

4. The method for rapid detection of a mobile terminal involved in a case based on signal perception according to claim 1, characterized in that, The random access preamble sequence after removing the cyclic prefix is ​​processed by Fast Fourier Transform, and the carrier frequency offset estimate and the inter-symbol phase difference sequence are extracted in the frequency domain, including: Perform a Fast Fourier Transform on the random access preamble sequence after removing the cyclic prefix to obtain a frequency domain sequence; Extract the frequency domain response values ​​of two subcarriers with a set index interval from the frequency domain sequence; The estimated carrier frequency offset is calculated based on the phase angle of the conjugate multiplication result of the frequency domain response values ​​of the two subcarriers; The frequency response values ​​corresponding to adjacent symbols in the frequency domain sequence are multiplied by conjugate, and the phase difference value of the conjugate multiplication result is extracted. The phase difference values ​​corresponding to all adjacent symbols in the random access preamble sequence are arranged in time order to generate the inter-symbol phase difference sequence.

5. A method for rapid detection of a mobile terminal involved in a case based on signal perception according to claim 1, characterized in that, The baseband physical layer feature vector is generated by concatenating the carrier frequency offset estimate and the inter-symbol phase difference sequence in a time sequence, including: The estimated carrier frequency offset is normalized and quantized to generate a frequency offset quantization sequence; The inter-symbol phase difference sequence is subjected to symbol bit value mapping processing to generate a phase difference quantization sequence; The frequency offset quantization sequence and the phase difference quantization sequence are alternately interpolated according to the timestamp order of the random access preamble sequence to generate an intermediate splicing sequence; The intermediate spliced ​​sequence is windowed and truncated to extract the portion of the intermediate spliced ​​sequence that corresponds to the valid data segment of the random access preamble sequence. The truncated sequence is then used as the baseband physical layer feature vector.

6. The method for rapid detection of a mobile terminal involved in a case based on signal perception according to claim 1, characterized in that, Perform a time-domain sliding cross-correlation operation between the baseband physical layer feature vector and a pre-stored baseband fingerprint database of the terminals involved in the case, including: Multiple reference feature vectors of the terminals involved in the case were retrieved from the baseband fingerprint database of the terminals involved in the case; The baseband physical layer feature vector is moved by a preset sliding step size, and the sum of the dot products of the moved baseband physical layer feature vector and each of the terminal reference feature vectors involved in the case is calculated to obtain multiple cross-correlation operation results. The maximum value among the multiple cross-correlation results is extracted as the cross-correlation peak value; If the cross-correlation peak value exceeds the preset decision threshold, the frequency offset change rate corresponding to the reference feature vector of the terminal involved in the case that generated the cross-correlation peak value is obtained, and it is determined whether the difference between the change rate of the carrier frequency offset estimate and the obtained frequency offset change rate is within a preset error range. If it is within the preset error range, the detection trigger signal is output.

7. A method for rapid detection of a mobile terminal involved in a case based on signal perception according to claim 2, characterized in that, The frequency of the local oscillator signal is dynamically configured according to the center frequency of the target frequency band, including: Scan the frequency band range corresponding to the target frequency band to obtain the signal power spectral density at each frequency point within the frequency band range; The frequency points in the signal power spectral density that are greater than a preset power threshold are identified as the main interference frequency points; Calculate the frequency difference between the center frequency of the target frequency band and the main interference frequency point, and adjust the control word of the numerically controlled oscillator according to the frequency difference so that the frequency of the adjusted local oscillator signal deviates from the main interference frequency point; Using the baseband IQ data stream after the subtraction compensation operation, cross-correlation-based time delay estimation is performed on the baseband IQ data streams corresponding to different channels in the multi-channel RF receiver, and time alignment operation is performed on the baseband IQ data streams of different channels according to the time delay estimation results.

8. A method for rapid detection of a mobile terminal involved in a case based on signal perception according to claim 3, characterized in that, After outputting the random access preamble sequence after removing the cyclic prefix, the method further includes: Extract the reference signal configuration parameters contained in the physical layer control signaling; The reference signal sequence is separated from the random access preamble sequence after removing the cyclic prefix according to the reference signal configuration parameters; Channel impulse response estimation is performed using the reference signal sequence to obtain the channel impulse response sequence; The number of multipath components and the arrival time of each multipath component are determined based on the power delay distribution of the channel impulse response sequence. When the number of multipath components exceeds a preset threshold, frequency domain equalization filter coefficients are constructed based on the channel impulse response sequence; The frequency domain equalization filter coefficients are used to perform a frequency domain linear equalization operation on the random access preamble sequence after removing the cyclic prefix.

9. A method for rapid detection of a mobile terminal involved in a case based on signal perception according to claim 4, characterized in that, After calculating the carrier frequency offset estimate based on the phase angle of the conjugate multiplication result of the frequency domain response values ​​of the two subcarriers, the method further includes: The estimated carrier frequency offset is processed by separating the fractional and integer multiples of frequency offset. Phase rotation compensation is performed on the frequency domain sequence using the separated fractional frequency offset; The frequency domain sequence after phase rotation compensation is subjected to inverse fast Fourier transform to obtain the time domain compensated sequence. Autocorrelation is performed on the first and last data segments of the time-domain compensation sequence to calculate the integer multiples of the frequency offset; The integer multiple frequency offset and the fractional multiple frequency offset are superimposed to generate a refined carrier frequency offset; The inter-symbol phase difference sequence is recalculated using the refined carrier frequency offset, and the phase jump points caused by noise in the recalculated inter-symbol phase difference sequence are smoothed by linear interpolation.

10. A method for rapid detection of a mobile terminal involved in a case based on signal perception according to claim 6, characterized in that, After outputting the detection trigger signal and locking the timestamp of the corresponding baseband data frame, the method further includes: Based on the locked timestamp, extract consecutive data frames from the baseband IQ data stream that are within a preset delay range after the timestamp; The continuous data frames are demodulated and channel decoded to extract Media Access Control (MAC) protocol data units. Parse the header field of the Media Access Control Layer Protocol data unit and extract the wireless network temporary identifier from the header field; The temporary wireless network identifier is hash-compared with a pre-stored database of identifiers for terminals involved in the case. If the hash comparison results are consistent, a final confirmation instruction is generated, and the final confirmation instruction is bound and stored with the locked timestamp.

Images