A mobile terminal quantity identification method based on uplink signal polarization state difference

Abstract

The application relates to a mobile terminal quantity identification method based on uplink signal polarization state difference, which is mainly used for quantity detection of mobile communication equipment with double-polarized antennas in a certain range. The method comprises the following steps: firstly, completing an air interface data acquisition task by using a data acquisition unit and buffering air interface IQ data; secondly, performing time-frequency analysis on the buffered data and restoring time-frequency domain information of the IQ data by a data processing unit through a sliding window method for fast Fourier transform; thirdly, separating part of time-frequency resource blocks and extracting antenna polarization information in different resource blocks by a feature extraction unit; and finally, obtaining equipment quantity information by clustering the extracted antenna polarization information by a quantity detection unit. The application can accurately obtain quantity information of mobile communication equipment in a certain range without any prior information.

Description

Technical Field

[0001] This invention relates to the field of mobile terminal quantity identification technology, and in particular to a method for identifying the number of mobile terminals based on the difference in uplink signal polarization state. Background Technology

[0002] In recent years, communication technology has seen significant improvements in network architecture, trust models, cryptographic algorithms, integrity protection, and user privacy protection. While numerous security strategies have been proposed to reduce the probability of communication systems being compromised, they still cannot avoid the biggest drawback of wireless communication: air interface data can be sniffed and analyzed by any radio receiving device.

[0003] Previous research has primarily relied on extracting device traffic fingerprints to detect the number of terminals. However, the biggest problem with fingerprint-based detection methods is fingerprint updating; offline learning and fingerprint extraction methods with relatively fixed patterns struggle to capture long-term stable device fingerprints. Furthermore, identifying the number of devices by extracting device fingerprints is overly complex. Number identification only requires capturing the differences in certain features between devices over a period of time, without needing to maintain the stability of these differences. Therefore, this application constructs features that can be distinguished within a short timeframe to detect and identify the number of mobile terminals within a certain range. Summary of the Invention

[0004] To avoid the shortcomings of the prior art, the present invention provides a mobile communication interference detection method based on time-frequency resource occupancy. This method can accurately detect the number of mobile terminals within the current detection range without any prior knowledge of the operator.

[0005] To achieve the objective of this invention, the technical solution adopted is: a method for identifying the number of mobile terminals based on the difference in uplink signal polarization state, comprising the following steps:

[0006] S1. Acquire mobile communication air interface IQ data and cache the IQ data within a predetermined time period;

[0007] S2. Perform time-frequency analysis on the cached IQ data to restore the time-frequency distribution status of the mobile communication physical resource blocks;

[0008] S3. Obtain the accurate physical resource block distribution location based on the time-frequency distribution state of the physical resource block and extract the polarization state near the location, and perform noise reduction processing.

[0009] S4. Use clustering algorithms to perform clustering learning on the extracted polarization state information, and obtain the number of clusters through clustering algorithms to complete the detection of the number of mobile communication devices within the detection range.

[0010] Preferably, the step of collecting IQ data in step S1 includes:

[0011] First, set the sampling center frequency, bandwidth, and sampling frequency, and then collect IQ data under the above parameters;

[0012] Select the truncation time to truncate the collected IQ data at the truncation time.

[0013] The required physical space is calculated based on the sampling frequency and time interval, and the IQ data is requested, stored, transmitted, and released. The IQ data is stored separately, divided into I-channel data and Q-channel data.

[0014] Preferably, step S2 includes the following steps:

[0015] Read the I-channel data and Q-channel data, and combine the I-channel data and Q-channel data into complex data;

[0016] Time-frequency analysis of complex data is performed using a time-domain sliding window or Fourier transform, and the size of a physical resource block in the time-frequency resource matrix is ​​calculated based on hyperparameters.

[0017] Coarse-grained filtering and noise reduction are applied to the time-frequency resource matrix to obtain the first-stage time-frequency resource matrix.

[0018] Preferably, step S3 includes the following steps:

[0019] Extract the exact data of the PUCCH from the time-frequency resource matrix based on the size of the physical resource block;

[0020] After finding the physical resource block location index, the physical resource block sent by the user terminal is extracted from the time-frequency resource matrix according to the coverage of the physical resource block.

[0021] Extract the polarization state of the captured physical resource block;

[0022] Noise filtering at the polarization state level is performed by determining the amplitude ratio range.

[0023] Preferably, the clustering algorithm includes, but is not limited to, the DBSCAN algorithm.

[0024] Preferably, a mobile terminal number identification system based on uplink signal polarization state differences includes a data acquisition unit, a data processing unit, a feature extraction unit, and a number detection unit;

[0025] The data acquisition unit is used to receive mobile communication air interface IQ data and cache IQ data within a predetermined time period for data processing.

[0026] The data processing unit is used to process the IQ data cached by the data acquisition unit within a predetermined time period, and to perform time-frequency analysis on the IQ data to restore the time-frequency distribution state of the mobile communication physical resource block.

[0027] The feature extraction unit is used to extract the distribution location of some physical resource blocks from the time-frequency analysis results, extract the physical resource block corresponding to the center point and extract its polarization state;

[0028] The number detection unit is used to perform clustering learning on the extracted polarization state information and obtain the number of clusters through a clustering algorithm.

[0029] Preferably, the data acquisition unit includes a hyperparameter determination module, an IQ data truncation module, and a physical resource request and storage module. The hyperparameter determination module is used to select hyperparameters, including sampling frequency, sampling center frequency, and bandwidth. The IQ data truncation module is used to select a time interval and truncate the IQ data acquired within the time interval. The physical resource request and storage module is used to store the IQ data.

[0030] Preferably, the data processing unit includes a cached IQ data reading module, an IQ data combination module, a combined data time-frequency analysis module, and a time-frequency filtering and noise reduction module. The cached IQ data reading module is used to read stored IQ data; the IQ data combination module is used to combine I-channel data and Q-channel data into complex data; the combined data time-frequency analysis module is used to select a time-frequency analysis method and hyperparameters; and the time-frequency filtering and noise reduction module performs a first-stage coarse-grained filtering based on the low-power and high-frequency transition characteristics of the noise signal.

[0031] Preferably, the feature extraction unit includes a time-frequency resource block location retrieval module, a time-frequency resource block truncation module, a polarization state extraction module for truncation data, and a noise filtering module. The time-frequency resource block location retrieval module is used to obtain the location of physical resource blocks; the time-frequency resource block truncation module is used to truncate physical resource blocks sent by the user terminal in the time-frequency resource matrix; the polarization state extraction module for truncation data is used to extract the polarization state within the physical resource block; and the noise filtering module is used to perform noise filtering processing at the polarization state level.

[0032] Preferably, the number detection unit includes a two-dimensional data clustering module and an outlier and outlier number processing module. The two-dimensional data clustering module is used to select hyperparameters for the scanning radius and minimum number of points. The outlier and outlier number processing module is used to identify and correct polarization state outliers and identification result anomalies.

[0033] The beneficial effects of the present invention are: (1) The present invention provides a novel approach to the identification of the number of mobile devices from the perspective of traffic analysis. This approach can complete the task of identifying the number of mobile communication devices within a certain range without cooperation or any prior information. Moreover, this implementation method avoids the disadvantages of high complexity and poor scalability of traditional device fingerprint extraction, and avoids the problems of decoding and demodulation of downlink control information and privacy leakage.

[0034] (2) This invention finds the differences in equipment by antenna polarization state. Antenna manufacturing process, mobile terminal placement position, mobile terminal movement, etc. will cause differences in polarization state. The DBSCAN algorithm is used to cluster the polarization state, which can effectively complete the clustering task of different polarization states.

[0035] (3) The method of the present invention has low data processing complexity, with computational complexity mainly concentrated in three parts: resource block location search, time-frequency analysis, and polarization state clustering. Resource block location search can reduce complexity by limiting the sampling bandwidth, and time-frequency analysis uses a sliding window fast Fourier algorithm. Attached Figure Description

[0036] Figure 1 This is a flowchart illustrating the method of the present invention;

[0037] Figure 2 This is a flowchart of the RB block extraction process during feature extraction;

[0038] Figure 3 This is a flowchart of polarization state extraction during the feature extraction process. Detailed Implementation

[0039] The present invention will now be described in further detail with reference to the accompanying drawings and preferred embodiments. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0040] As an optional implementation method in this embodiment, such as Figure 1 As shown, a mobile terminal quantity identification system based on uplink signal polarization state differences includes a data acquisition unit, a data processing unit, a feature extraction unit, and a quantity detection unit. The data acquisition unit receives mobile communication air interface IQ data and caches IQ data within a predetermined time period for data processing. The data processing unit processes the IQ data cached by the data acquisition unit within the predetermined time period and performs time-frequency analysis on the IQ data to recover the time-frequency distribution state of mobile communication physical resource blocks. The feature extraction unit extracts the distribution positions of some physical resource blocks from the time-frequency analysis results, extracts the physical resource block corresponding to the center point, and extracts its polarization state. The quantity detection unit performs clustering learning on the extracted polarization state information and obtains the number of clusters through a clustering algorithm.

[0041] Specifically, the data acquisition unit includes a hyperparameter determination module, an IQ data truncation module, and a physical resource request and storage module. The hyperparameter determination module is used to select hyperparameters, including sampling frequency, sampling center frequency, and bandwidth. The IQ data truncation module is used to select time intervals and truncate the IQ data acquired within the time intervals. The physical resource request and storage module is used to store the IQ data.

[0042] The data processing unit includes a cached IQ data reading module, an IQ data combination module, a combined data time-frequency analysis module, and a time-frequency filtering and noise reduction module. The cached IQ data reading module is used to read the stored IQ data; the IQ data combination module is used to combine the I-channel data and Q-channel data into complex data; the combined data time-frequency analysis module is used to select the time-frequency analysis method and hyperparameters; and the time-frequency filtering and noise reduction module performs coarse-grained filtering in the first stage based on the low-power and high-frequency transition characteristics of the noise signal.

[0043] The feature extraction unit includes a time-frequency resource block location retrieval module, a time-frequency resource block truncation module, a polarization state extraction module for truncation data, and a noise filtering module. The time-frequency resource block location retrieval module is used to obtain the location of physical resource blocks; the time-frequency resource block truncation module is used to truncate physical resource blocks sent by user terminals in the time-frequency resource matrix; the polarization state extraction module for truncation data is used to extract the polarization state within the physical resource block; and the noise filtering module is used for noise filtering at the polarization state level.

[0044] The number detection unit includes a two-dimensional data clustering module and an outlier and outlier number processing module. The two-dimensional data clustering module is used to select hyperparameters for the scanning radius and minimum number of points; the outlier and outlier number processing module is used to identify and correct polarization state outliers and anomalies in the recognition results.

[0045] Based on the above system, this invention discloses a method for identifying the number of mobile terminals based on the difference in uplink signal polarization state, the method comprising the following steps:

[0046] First, mobile communication air interface IQ data is acquired and cached within a predetermined time period. Specifically, for FDD standard, uplink PUCCH channel signals are collected. Hyperparameters are selected first through a hyperparameter determination module; these hyperparameters mainly include sampling frequency, bandwidth, and sampling center frequency. Then, the acquired IQ data is truncated by an IQ data truncation module. In this embodiment, a 100ms time interval is selected, and truncation occurs at 100ms of the acquired IQ data. Finally, the IQ data is stored through a physical resource request and storage module. In this embodiment, IQ values ​​are stored in float32 format and divided into two data channels: one for I-channel data and the other for Q-channel data. The physical space required for the IQ data is calculated based on the sampling frequency and time interval, and the IQ data can be requested, stored, transmitted, and released.

[0047] Next, time-frequency analysis is performed on the cached IQ data to restore the time-frequency distribution state of the mobile communication physical resource blocks. Specifically, the cached IQ data is first read by the cached IQ data reading module in float32 format. Then, the IQ data and Q data are combined into complex data by the IQ data combination module. Next, the combined data time-frequency analysis module performs time-frequency analysis on the complex data. Specifically, this module includes the selection of time-frequency analysis methods and hyperparameters. In this embodiment, the time-frequency analysis method is short-time Fourier transform, which is equivalent to multiplying the data within the sliding window by a window function and performing a fast Fourier transform. In this embodiment, the required hyperparameters are sampling frequency, number of fast Fourier transform points, window function, and window movement step size. The exact size of an RB block (physical resource block) in the time-frequency matrix can be calculated based on the number of fast Fourier transform points and the window movement step size. Finally, filtering and noise reduction are performed by the time-frequency filtering and noise reduction module. This filtering and noise reduction is a first-stage coarse-grained filtering, which filters out noise signals based on their low power and high-frequency jump characteristics.

[0048] Then, based on the time-frequency distribution of physical resource blocks, the accurate distribution location of the physical resource blocks is obtained, and the polarization state near that location is extracted, followed by noise reduction processing. Specifically, the time-frequency resource matrix obtained from the data processing unit is further processed to obtain the accurate distribution location of the physical resource blocks and extract the polarization state near that location. Specifically, a time-frequency resource block location retrieval module based on a sliding window averaging is first used; the process of this module can be found in [reference needed]. Figure 2That is, firstly, the exact data of the PUCCH is extracted from the time-frequency resource matrix. The extraction is based on the exact size of the RB block calculated in the time-frequency analysis. These steps effectively reduce the amount of invalid background data processed and improve computational efficiency. Due to the frequency leakage phenomenon of the Fast Fourier Transform in actual implementation, neighboring time-frequency information is needed for background reduction. Next, the average power within the window is calculated using a sliding window based on the RB size. An average power curve is plotted, and mean filtering is applied to the average power curve to find the peak value. The index corresponding to the peak value is then found, and the peak index is the location of the sought RB block. Then, the RB block is extracted using the time-frequency resource block extraction module. Specifically, after finding the RB block location index, the RB block sent by the user terminal can be extracted from the time-frequency resource matrix based on the RB block coverage area. Finally, the polarization state within the RB block is extracted using the polarization state extraction module. The extraction process for the polarization state within the RB block can be found in [reference needed]. Figure 3 Specifically, within the RB blocks extracted from the dual-channel data, the amplitude and phase spectra in the frequency domain are calculated respectively. The polarization state can be measured by both the amplitude ratio and the phase difference. Since the phase difference distribution ranges from [0, 2π], it can be visualized using a polar coordinate scatter plot. Finally, a two-dimensional vector of amplitude ratio and phase difference is constructed for subsequent numerical clustering. Finally, a noise filtering module based on polarization state is used for filtering. Specifically, the polarization state can be measured by both the amplitude ratio and the phase difference. In RB blocks with user data transmission, the amplitudes of the two channels should be similar, but noise does not meet this condition. Noise filtering at the polarization state level can be performed by judging the amplitude ratio range.

[0049] Finally, a clustering algorithm is used to perform clustering learning on the extracted polarization state information, and the number of clusters is obtained through the clustering algorithm to complete the detection of the number of mobile communication devices within the detection range. Specifically, the extracted polarization state is used to identify the number of mobile terminals within the collection period. Specifically, a two-dimensional data clustering module based on the DBSCAN algorithm is used to perform cluster analysis on the extracted polarization state information, involving the selection of hyperparameters for the scanning radius and minimum number of points. These two hyperparameters limit the density of each cluster in the DBSCAN algorithm. Compared to the K-means algorithm, this algorithm does not require providing the number of clusters. Finally, an outlier and outlier number processing module is used. This module includes the identification and correction of polarization state outliers and recognition result anomalies. The correction of outliers mainly focuses on the selection range of normal amplitude ratios, and the correction of recognition result anomalies mainly addresses the situation where no users are present and only noise is clustered.

[0050] The above description is only a specific embodiment of the present invention. Various examples and illustrations do not constitute a limitation on the substantive content of the present invention. Those skilled in the art can make modifications or variations to the above-described specific embodiments after reading the specification without departing from the substance and scope of the invention.

Claims

1. A method for identifying the number of mobile terminals based on differences in uplink signal polarization states, characterized in that, Includes the following steps: S1. Acquire mobile communication air interface IQ data and cache the IQ data within a predetermined time period. The steps for acquiring IQ data include: First, set the sampling center frequency, bandwidth, and sampling frequency, and then collect IQ data under the above parameters; Select the truncation time to truncate the collected IQ data at the truncation time. The required physical space is calculated based on the sampling frequency and time interval, and the IQ data is requested, stored, transmitted and released. The IQ data is stored separately, divided into I-channel data and Q-channel data. S2. Perform time-frequency analysis on the cached IQ data to restore the time-frequency distribution status of mobile communication physical resource blocks, including the following steps: Read the I-channel data and Q-channel data, and combine the I-channel data and Q-channel data into complex data; Time-frequency analysis of complex data is performed using a time-domain sliding window or Fourier transform, and the size of a physical resource block in the time-frequency resource matrix is ​​calculated based on hyperparameters. Coarse-grained filtering and noise reduction are applied to the time-frequency resource matrix to obtain the first-stage time-frequency resource matrix; S3. Obtain the accurate physical resource block distribution location based on the time-frequency distribution state of the physical resource block and extract the polarization state near that location, and perform noise reduction processing, including the following steps: Extract the exact data of the PUCCH from the time-frequency resource matrix based on the size of the physical resource block; After finding the physical resource block location index, the physical resource block sent by the user terminal is extracted from the time-frequency resource matrix according to the coverage of the physical resource block. Extract the polarization state of the captured physical resource block; Noise filtering at the polarization state level is performed by determining the amplitude ratio range; S4. Use clustering algorithms to perform clustering learning on the extracted polarization state information, and obtain the number of clusters through clustering algorithms to complete the detection of the number of mobile communication devices within the detection range.

2. The method for identifying the number of mobile terminals based on the difference in uplink signal polarization state according to claim 1, characterized in that: The clustering algorithm includes the DBSCAN algorithm.

3. A mobile terminal quantity identification system based on uplink signal polarization state difference, applied to the mobile terminal quantity identification method based on uplink signal polarization state difference as described in any one of claims 1-2, characterized in that: It includes a data acquisition unit, a data processing unit, a feature extraction unit, and a number detection unit; The data acquisition unit is used to receive mobile communication air interface IQ data and cache IQ data within a predetermined time period for data processing. The data processing unit is used to process the IQ data cached by the data acquisition unit within a predetermined time period, and to perform time-frequency analysis on the IQ data to restore the time-frequency distribution state of the mobile communication physical resource block. The feature extraction unit is used to extract the distribution location of some physical resource blocks from the time-frequency analysis results, extract the physical resource block corresponding to the center point and extract its polarization state; The number detection unit is used to perform clustering learning on the extracted polarization state information and obtain the number of clusters through a clustering algorithm.

4. The mobile terminal quantity identification system based on uplink signal polarization state difference according to claim 3, characterized in that: The data acquisition unit includes a hyperparameter determination module, an IQ data truncation module, and a physical resource request and storage module. The hyperparameter determination module is used to select hyperparameters, including sampling frequency, sampling center frequency, and bandwidth. The IQ data truncation module is used to select a time interval and truncate the IQ data acquired within the time interval. The physical resource request and storage module is used to store the IQ data.

5. The mobile terminal quantity identification system based on uplink signal polarization state difference according to claim 4, characterized in that: The data processing unit includes a cached IQ data reading module, an IQ data combination module, a combined data time-frequency analysis module, and a time-frequency filtering and noise reduction module. The cached IQ data reading module is used to read stored IQ data; the IQ data combination module is used to combine I-channel data and Q-channel data into complex data; the combined data time-frequency analysis module is used to select time-frequency analysis methods and hyperparameters; and the time-frequency filtering and noise reduction module performs coarse-grained filtering in the first stage based on the low-power and high-frequency transition characteristics of the noise signal.

6. The mobile terminal quantity identification system based on uplink signal polarization state difference according to claim 5, characterized in that: The feature extraction unit includes a time-frequency resource block location retrieval module, a time-frequency resource block truncation module, a polarization state extraction module for truncation data, and a noise filtering module. The time-frequency resource block location retrieval module is used to obtain the location of physical resource blocks; the time-frequency resource block truncation module is used to truncate physical resource blocks sent by user terminals in the time-frequency resource matrix; the polarization state extraction module for truncation data is used to extract the polarization state within the physical resource block; and the noise filtering module is used to perform noise filtering processing at the polarization state level.

7. The mobile terminal quantity identification system based on uplink signal polarization state difference according to claim 6, characterized in that: The number detection unit includes a two-dimensional data clustering module and an outlier and outlier number processing module. The two-dimensional data clustering module is used to select hyperparameters for the scanning radius and minimum number of points. The outlier and outlier number processing module is used to identify and correct polarization state outliers and identification result anomalies.

Images