GNSS interference anomaly detection and identification system and method

Abstract

The invention relates to the technical field of satellite navigation, in particular to a GNSS interference anomaly detection and recognition system and method. Comprising a multi-system multi-frequency-point receiver module, an ionized layer scintillation identification module, a solar radio identification module, a satellite signal abnormity identification module, a time-frequency domain analysis module, an electromagnetic interference detection module, an interference characteristic database management module and an electromagnetic interference intelligent identification matching module. The detection precision is improved; secondly, the method has a quick response capability, can quickly give an alarm when interference occurs, and reduces adverse effects on services; in addition, the interference type can be identified, clear interference information is provided for the user, and the response flexibility is enhanced; and finally, the data storage and backtracking functions are convenient for subsequent deep analysis and system optimization, and data support and guarantee are provided for service development, so that the problem that GNSS interference signals are difficult to effectively detect in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of satellite navigation technology, in particular to a GNSS interference anomaly detection and identification system and method. BACKGROUND

[0002] The development of multi-constellation and multi-frequency in global navigation satellite system (GNSS) significantly improves the positioning accuracy and reliability, and promotes the application of advanced receiver autonomous integrity monitoring (ARAIM). ARAIM effectively reduces the vertical protection level (VPL) by fusing multi-source observation information and probability model, and supports precise vertical guidance (such as LPV-200), and the core is to detect satellite faults (such as ephemeris error) by comparing the "consensus" between satellite signals.

[0003] However, ARAIM and its predecessor RAIM, which are designed to deal with internal random faults, are helpless against the increasingly serious external malicious interference. For example: suppressing interference submerges weak GNSS signals, leading to loss of lock; deceptive interference transmits a realistic false signal to induce the receiver to output dangerous error information. The low power and dynamic characteristics of modern interference, and the premise assumption of ARAIM that depends on "most satellite integrity" is destroyed by interference, making it difficult for traditional methods to effectively detect, especially for carefully designed deceptive attacks. SUMMARY

[0004] The purpose of the present application is to provide a GNSS interference anomaly detection and identification system and method, which aims to solve the problem that it is difficult to effectively detect GNSS interference signals in the prior art.

[0005] To achieve the above purpose, in a first aspect, the present application provides a GNSS interference anomaly detection and identification system, comprising a multi-system multi-frequency receiver module, an ionospheric scintillation identification module, a solar radio identification module, a satellite signal anomaly identification module, a time-frequency domain analysis module, an electromagnetic interference detection module, an interference feature database management module and an electromagnetic interference intelligent identification matching module. The ionospheric scintillation identification module, the solar radio identification module, the satellite signal anomaly identification module, the time-frequency domain analysis module and the electromagnetic interference detection module are connected with the multi-system multi-frequency receiver module, and the electromagnetic interference intelligent identification matching module is connected with the interference feature database management module and the electromagnetic interference detection module.

[0006] The multi-system multi-frequency receiver module is used for receiving GNSS intermediate frequency signals, performing acquisition, bit synchronization, tracking, and text decoding operations, and outputting carrier-to-noise ratio, Doppler, carrier phase, pseudorange, and positioning results.

[0007] The ionospheric scintillation identification module is configured to identify whether GNSS signals are affected by ionospheric scintillation and estimate ionospheric scintillation characteristics according to signal processing parameters output by the multi-system multi-frequency receiver module.

[0008] The solar radio burst identification module is configured to identify whether GNSS signals are affected by solar radio bursts according to signal processing parameters output by the multi-system multi-frequency receiver module.

[0009] The satellite signal anomaly identification module is configured to determine whether satellite signals are abnormal according to signal processing parameters output by the multi-system multi-frequency receiver module.

[0010] The time-frequency domain analysis module is configured to perform time-frequency domain analysis on received GNSS intermediate frequency signals to extract time-frequency domain characteristic parameters of electromagnetic interference signals.

[0011] The electromagnetic interference detection module is configured to determine whether electromagnetic interference exists in signals according to signal processing parameters output by the multi-system multi-frequency receiver module, and to call the electromagnetic interference intelligent matching identification module according to the detection result.

[0012] The interference characteristic database management module is configured to store characteristic sample data of different types of interference signals in multiple scenarios, and to transmit interference characteristic training data to the electromagnetic interference identification module according to a received intermediate frequency signal scenario.

[0013] The electromagnetic interference intelligent identification matching module is configured to train and learn data samples of different types of electromagnetic interference in a given scenario, and to classify the to-be-tested samples composed of characteristic parameters according to a training function.

[0014] The multi-system multi-frequency receiver module includes a signal acquisition unit, a bit synchronization unit, a signal tracking unit, and a positioning solution unit, and the signal acquisition unit, the bit synchronization unit, the signal tracking unit, and the positioning solution unit are connected in sequence.

[0015] The signal acquisition unit is configured to quickly estimate Doppler frequency offset and initial code phase to complete signal acquisition.

[0016] The bit synchronization unit is configured to lock the bit transition boundary of navigation messages by histogram statistics.

[0017] The signal tracking unit is configured to jointly track the carrier and code phase by using a phase-locked loop and a delay-locked loop to output carrier-to-noise ratio, Doppler frequency shift, carrier phase, and pseudo-range in real time.

[0018] The positioning solution unit is configured to parse ephemeris and clock correction data in navigation messages, fuse multi-satellite pseudo-ranges by using a least squares method or Kalman filtering, and output position, velocity, and time results.

[0019] The ionospheric scintillation identification module comprises a parameter analysis unit, a feature extraction unit and a dynamic decision unit, and the parameter analysis unit, the feature extraction unit and the dynamic decision unit are sequentially connected.

[0020] The parameter analysis unit is configured to analyze the carrier-to-noise ratio, the carrier phase error and the phase-locked loop jitter parameter output by the receiver.

[0021] The feature extraction unit is configured to calculate the amplitude scintillation index and the phase scintillation index.

[0022] The dynamic decision unit is configured to dynamically adjust the threshold in combination with the geographic latitude and the Kp index, and output the scintillation intensity grade, the occurrence time and the duration.

[0023] The solar radio identification module comprises a carrier-to-noise ratio monitoring unit, a time matching unit and an interference confirmation unit, and the carrier-to-noise ratio monitoring unit, the time matching unit and the interference confirmation unit are sequentially connected.

[0024] The carrier-to-noise ratio monitoring unit is configured to detect whether the carrier-to-noise ratios of multiple frequency points appear synchronism and non-locality sudden drop.

[0025] The time matching unit is configured to match the interference occurrence time with the solar flare burst time.

[0026] The interference confirmation unit is configured to exclude local human interference by the similarity of multiple frequency points affected by interference, mark the affected satellite and reduce the weight of the positioning result.

[0027] The satellite signal anomaly identification module comprises a code carrier separation detection unit, a signal distortion judgment unit, a data verification unit and a fault positioning unit, and the code carrier separation detection unit, the signal distortion judgment unit and the data verification unit are respectively connected with the fault positioning unit.

[0028] The code carrier separation detection unit is configured to monitor the difference between the pseudorange and the carrier phase change rate.

[0029] The signal distortion judgment unit is configured to detect the correlation peak asymmetry by using a plurality of correlators.

[0030] The data verification unit is configured to verify the navigation message CRC and the reasonableness of ephemeris parameters.

[0031] The fault positioning unit is configured to position the single satellite fault by using geometric distance residual analysis, and output the abnormal satellite PRN number and type.

[0032] The time-frequency domain analysis module comprises a time domain analysis unit, a frequency domain analysis unit, a time-frequency joint analysis unit and a parameter output unit, and the time domain analysis unit, the frequency domain analysis unit and the time-frequency joint analysis unit are connected with the parameter output unit respectively.

[0033] The time domain analysis unit is configured to calculate an amplitude histogram to identify impulse interference and detect periodic interference through an autocorrelation function.

[0034] The frequency domain analysis unit is configured to adopt FFT spectrum analysis to locate narrowband / single-frequency interference and utilize spectral entropy values to reduce energy-concentrated interference.

[0035] The time-frequency joint analysis unit is configured to determine an interference time period and a frequency band through a short-time Fourier transform and capture transient Chirp signal characteristics in combination with a wavelet transform.

[0036] The parameter output unit is configured to output interference bandwidth, center frequency and modulation type parameters.

[0037] The electromagnetic interference detection module comprises an index monitoring unit, a fusion decision unit and a trigger control unit, and the index monitoring unit, the fusion decision unit and the trigger control unit are connected in sequence.

[0038] The index monitoring unit is configured to monitor, in real time, sudden drop of carrier-to-noise ratio, distortion of correlation peak and sudden increase of tracking loop loss rate.

[0039] The fusion decision unit is configured to adaptively fuse indexes by using a logistic regression model and output a detection confidence.

[0040] The trigger control unit is configured to automatically call an electromagnetic interference intelligent identification module when the confidence is greater than 90% and output a preliminary interference classification.

[0041] The interference feature database management module comprises a hierarchical storage unit, a scene matching unit, an incremental learning unit and a data stream output unit, and the incremental learning unit, the hierarchical storage unit, the scene matching unit and the data stream output unit are connected in sequence.

[0042] The hierarchical storage unit is configured to store sample features and scene labels according to interference types.

[0043] The scene matching unit is configured to automatically load a matching scene library according to a receiver geographic position.

[0044] The incremental learning unit is configured to support an administrator to add new samples and update feature indexes through an interface.

[0045] The data stream output unit is configured to provide a training sample set to an identification module in a standardized JSON format.

[0046] The electromagnetic interference intelligent identification matching module comprises a training unit, a classification identification unit, a sample matching unit and a result output unit, the training unit, the classification identification unit and the result output unit are connected in sequence, and the sample matching unit is connected with the result output unit;

[0047] The training unit is configured to train a classifier model using sample data in an interference feature database.

[0048] The classification identification unit is configured to input a feature vector extracted by the time-frequency domain module into a classifier to obtain a probabilistic interference type label.

[0049] The sample matching unit is configured to match the most similar historical sample in the database based on a K-nearest neighbor algorithm.

[0050] The result output unit is configured to output an interference type, a confidence degree and a matching sample ID, and drive an anti-interference strategy generation.

[0051] In a second aspect, a GNSS interference anomaly detection and identification method is used in the GNSS interference anomaly detection and identification system of the first aspect, and comprises the following steps:

[0052] Receiving and resolving GNSS intermediate frequency signals to output basic observation parameters;

[0053] Detecting the influence of ionospheric scintillation and solar radio burst on signals;

[0054] Identifying satellite abnormalities;

[0055] Multi-index fusion judgment of electromagnetic interference existence;

[0056] Time domain / frequency domain / time-frequency joint analysis of interference characteristics;

[0057] Machine learning classification of interference types and matching of historical samples;

[0058] Activation of ARAIM algorithm to reduce VPL when satellite geometry deteriorates;

[0059] Output of a suppression scheme based on all detection results.

[0060] The GNSS interference anomaly detection and identification system provided by the application comprises a multi-system multi-frequency point receiver module, an ionospheric scintillation identification module, a solar radio identification module, a satellite signal anomaly identification module, a time-frequency domain analysis module, an electromagnetic interference detection module, an interference feature database management module and an electromagnetic interference intelligent identification matching module. BRIEF DESCRIPTION OF DRAWINGS

[0061] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0062] Figure 1 FIG. 1 is a schematic diagram of a GNSS interference anomaly detection and identification system provided by the present application.

[0063] Figure 2 FIG. 2 is a schematic diagram of a multi-system multi-frequency point receiver module.

[0064] Figure 3 FIG. 3 is a schematic diagram of an ionospheric scintillation identification module.

[0065] Figure 4 FIG. 4 is a schematic diagram of a solar radio identification module.

[0066] Figure 5 FIG. 5 is a schematic diagram of a satellite signal anomaly identification module.

[0067] Figure 6 FIG. 6 is a schematic diagram of a time-frequency domain analysis module.

[0068] Figure 7 FIG. 7 is a schematic diagram of an electromagnetic interference detection module.

[0069] Figure 8is a schematic diagram of the interference feature database management module.

[0070] Figure 9 is a schematic diagram of the electromagnetic interference intelligent identification matching module.

[0071] Figure 10 is a flow chart of the GNSS interference anomaly detection and identification method provided by the present application.

[0072] In the figure: 1 - multi-system multi-frequency receiver module, 2 - ionospheric scintillation identification module, 3 - solar radio identification module, 4 - satellite signal anomaly identification module, 5 - time-frequency domain analysis module, 6 - electromagnetic interference detection module, 7 - interference feature database management module, 8 - electromagnetic interference intelligent identification matching module, 11 - signal acquisition unit, 12 - bit synchronization unit, 13 - signal tracking unit, 14 - positioning solution unit, 21 - parameter analysis unit, 22 - feature extraction unit, 23 - dynamic decision unit, 31 - carrier-to-noise ratio monitoring unit, 32 - time matching unit, 33 - interference confirmation unit, 41 - code carrier separation detection unit, 42 - signal distortion judgment unit, 43 - data verification unit, 44 - fault positioning unit, 51 - time domain analysis unit, 52 - frequency domain analysis unit, 53 - time-frequency joint analysis unit, 54 - parameter output unit, 61 - index monitoring unit, 62 - fusion decision unit, 63 - trigger control unit, 71 - hierarchical storage unit, 72 - scene matching unit, 73 - incremental learning unit, 74 - data stream output unit, 81 - training unit, 82 - classification identification unit, 83 - sample matching unit, 84 - result output unit. DETAILED DESCRIPTION

[0073] Embodiments of the present application are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by reference to the accompanying drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application.

[0074] Please refer to Figures 1 to 9In the first aspect, the application provides a GNSS interference anomaly detection and identification system, which comprises a multi-system multi-frequency receiver module 1, an ionospheric scintillation identification module 2, a solar radio identification module 3, a satellite signal anomaly identification module 4, a time-frequency domain analysis module 5, an electromagnetic interference detection module 6, an interference feature database management module 7 and an electromagnetic interference intelligent identification matching module 8. The ionospheric scintillation identification module 2, the solar radio identification module 3, the satellite signal anomaly identification module 4, the time-frequency domain analysis module 5 and the electromagnetic interference detection module 6 are connected with the multi-system multi-frequency receiver module 1 respectively. The electromagnetic interference intelligent identification matching module 8 is connected with the interference feature database management module 7 and the electromagnetic interference detection module 6.

[0075] The multi-system multi-frequency receiver module 1 is used for receiving GNSS intermediate frequency signals, performing acquisition, bit synchronization, tracking, and text decoding operations, and outputting carrier-to-noise ratio, Doppler, carrier phase, pseudo-range, and positioning results.

[0076] The ionospheric scintillation identification module 2 is used for identifying whether the GNSS signals are affected by ionospheric scintillation according to the signal processing parameters output by the multi-system multi-frequency receiver module 1 and giving an estimate of the ionospheric scintillation characteristic quantity.

[0077] The solar radio identification module 3 is used for identifying whether the GNSS signals are affected by solar radio bursts according to the signal processing parameters output by the multi-system multi-frequency receiver module 1.

[0078] The satellite signal anomaly identification module 4 is used for judging whether the satellite signal itself has an abnormal quality according to the signal processing parameters output by the multi-system multi-frequency receiver module 1.

[0079] The time-frequency domain analysis module 5 is used for performing time-frequency domain analysis and processing on the received GNSS intermediate frequency signals and extracting time-frequency domain characteristic parameters of the electromagnetic interference signals.

[0080] The electromagnetic interference detection module 6 is used for judging whether there is electromagnetic interference in the signals according to the signal processing parameters output by the multi-system multi-frequency receiver module 1 and calling the electromagnetic interference intelligent matching identification module according to the detection result.

[0081] The interference feature database management module 7 is used for storing characteristic sample data of different types of interference signals in multiple scenarios and transmitting interference feature training data to the electromagnetic interference identification module according to the received intermediate frequency signal scenario.

[0082] The electromagnetic interference intelligent identification matching module 8 is used for training and learning data samples of different types of electromagnetic interference in a given scenario and classifying the to-be-tested samples composed of characteristic parameters according to a training function.

[0083] In the embodiment, through the accurate extraction and multi-dimensional monitoring of signal characteristics by the feature analysis module, and the analysis and comparison by the interference identification module using advanced algorithms, various interference signals can be more accurately detected, the misjudgment rate and the missed judgment rate are reduced, and the detection accuracy is significantly improved compared with the prior art. From signal reception, feature analysis to interference identification and alarm, the whole system design focuses on fast response process. Once interference occurs, the alarm can be sent quickly, so that the user can know the interference situation in time and take measures to reduce the adverse effect of interference on GNSS application business. Not only can it detect whether interference exists, but also can further identify the specific type of interference, provide key basis for users to take more targeted solutions for different types of interference, enhance the flexibility and effectiveness of interference response, and thus solve the problem of difficult effective detection of GNSS interference signals in the prior art.

[0084] Further, the multi-system multi-frequency receiver module 1 comprises a signal acquisition unit 11, a bit synchronization unit 12, a signal tracking unit 13 and a positioning solution unit 14, which are connected in sequence.

[0085] The signal acquisition unit 11 is used for fast estimation of Doppler frequency offset and initial code phase, and completes signal acquisition.

[0086] The bit synchronization unit 12 is used for locking the bit transition boundary of navigation message by histogram statistics method.

[0087] The signal tracking unit 13 is used for joint phase-locked loop and delay-locked loop dynamic tracking of carrier and code phase, and real-time output of carrier-to-noise ratio, Doppler frequency shift, carrier phase and pseudo-range.

[0088] The positioning solution unit 14 is used for parsing ephemeris and clock correction data in navigation message, fusing multi-satellite pseudo-range by least square method or Kalman filter, and outputting position, velocity and time results.

[0089] In the embodiment, the multi-system multi-frequency receiver module 1 is responsible for real-time processing of multi-frequency GNSS intermediate frequency signals, and realizes the whole process solution from signal acquisition to positioning output through a software radio architecture. The Doppler frequency offset and initial code phase of the signal are quickly estimated to complete acquisition, and then bit synchronization is performed, and the bit transition boundary of the navigation message is locked through a histogram statistical method. After entering the tracking stage, the phase-locked loop (PLL) and the delay-locked loop (DLL) dynamically track the carrier and code phase, and real-time output key parameters such as carrier-to-noise ratio (C / N0), Doppler frequency shift, carrier phase and pseudo-range, and based on the tracking results, data such as ephemeris and clock correction in the navigation message are parsed, and finally the least squares method or Kalman filtering algorithm is used to fuse multi-satellite pseudo-range, and the position, velocity and time (PVT) results are solved.

[0090] Further, the ionospheric scintillation identification module 2 includes a parameter analysis unit 21, a feature extraction unit 22 and a dynamic decision unit 23, and the parameter analysis unit 21, the feature extraction unit 22 and the dynamic decision unit 23 are connected in sequence.

[0091] The parameter analysis unit 21 is configured to analyze the carrier-to-noise ratio, carrier phase error and phase-locked loop jitter parameters output by the receiver.

[0092] The feature extraction unit 22 is configured to calculate the amplitude scintillation index and the phase scintillation index.

[0093] The dynamic decision unit 23 is configured to dynamically adjust the threshold value in combination with the geographic latitude and the Kp index, and output the scintillation intensity level, occurrence time and duration.

[0094] In the embodiment, the ionospheric scintillation identification module 2 analyzes the carrier-to-noise ratio, carrier phase error and phase-locked loop jitter parameters output by the receiver, calculates key feature quantities to evaluate the positioning stability. Its working principle focuses on extracting the amplitude scintillation index and the phase scintillation index, dynamically adjusting the threshold value in combination with the geographic latitude and the Kp index, and finally outputting the scintillation intensity level, occurrence time and duration.

[0095] Further, the solar radio identification module 3 includes a carrier-to-noise ratio monitoring unit 31, a time matching unit 32 and an interference confirmation unit 33, and the carrier-to-noise ratio monitoring unit 31, the time matching unit 32 and the interference confirmation unit 33 are connected in sequence.

[0096] The carrier-to-noise ratio monitoring unit 31 is configured to detect whether the multi-frequency carrier-to-noise ratio appears synchronously and non-locally.

[0097] The time matching unit 32 is configured to match the interference occurrence time with the solar flare burst time.

[0098] The interference confirmation unit 33 is configured to exclude local human interference by multi-point interference similarity, mark affected satellites and reduce the weight of positioning results.

[0099] In the embodiment, the solar radio identification module 3 identifies the broadband electromagnetic noise interference caused by solar radio burst (SRB), monitors whether the multi-point carrier-to-noise ratio of the receiver output appears synchronous, significant and non-local sudden drop (such as >10 dB-Hz), matches the time of interference occurrence with the time of solar flare burst, excludes local human interference source by analyzing the similarity of multi-point interference degree, confirms, marks the affected satellites and the interference strength, and reduces the credibility of the related positioning results.

[0100] Further, the satellite signal anomaly identification module 4 includes a code carrier separation detection unit 41, a signal distortion judgment unit 42, a data verification unit 43 and a fault positioning unit 44, and the code carrier separation detection unit 41, the signal distortion judgment unit 42 and the data verification unit 43 are respectively connected with the fault positioning unit 44.

[0101] The code carrier separation detection unit 41 is configured to monitor the difference between pseudorange and carrier phase rate of change.

[0102] The signal distortion judgment unit 42 is configured to detect correlation peak asymmetry by using a multi-correlator.

[0103] The data verification unit 43 is configured to verify the navigation message CRC and the rationality of ephemeris parameters.

[0104] The fault positioning unit 44 is configured to locate single satellite fault by geometric distance residual analysis, and output the abnormal satellite PRN number and type.

[0105] In the embodiment, the satellite signal anomaly identification module 4 identifies code carrier separation (CCD) anomaly by monitoring the difference between pseudorange and carrier phase rate of change, judges signal distortion by using a multi-correlator to detect correlation peak asymmetry, finds data anomaly by verifying the navigation message CRC and the rationality of ephemeris parameters, and performs multi-satellite consistency test by geometric distance residual analysis to locate single satellite fault. Finally, the abnormal satellite PRN number and specific abnormal type are output to guide the receiver to exclude the fault satellite in positioning solution.

[0106] Further, the time-frequency domain analysis module 5 includes a time domain analysis unit 51, a frequency domain analysis unit 52, a time-frequency joint analysis unit 53 and a parameter output unit 54, and the time domain analysis unit 51, the frequency domain analysis unit 52 and the time-frequency joint analysis unit 53 are respectively connected with the parameter output unit 54.

[0107] The time domain analysis unit 51 is configured to calculate an amplitude histogram to identify impulse interference, and detect periodic interference through an autocorrelation function.

[0108] The frequency domain analysis unit 52 is configured to use FFT spectrum analysis to locate narrowband / single frequency interference, and use a spectral entropy value to reduce energy concentrated interference.

[0109] The time-frequency joint analysis unit 53 is configured to determine an interference time period and frequency band through a short-time Fourier transform, and capture transient Chirp signal characteristics through a wavelet transform.

[0110] The parameter output unit 54 is configured to output interference bandwidth, center frequency and modulation type parameters.

[0111] In the embodiment, the time-frequency domain analysis module 5 calculates an amplitude histogram to identify impulse interference at a time domain level, detects periodic interference through an autocorrelation function, uses FFT spectrum analysis to locate narrowband / single frequency interference at a frequency domain level, uses a spectral entropy value to reduce energy concentrated interference, determines an interference time period and frequency band through a short-time Fourier transform (STFT) at a time-frequency joint level, and captures transient Chirp signal characteristics through a wavelet transform. Finally, key parameters such as interference bandwidth, center frequency and modulation type are output.

[0112] Further, the electromagnetic interference detection module 6 includes an index monitoring unit 61, a fusion decision unit 62 and a trigger control unit 63, which are connected in sequence.

[0113] The index monitoring unit 61 is configured to monitor sudden decrease of carrier-to-noise ratio, correlation peak distortion and sudden increase of tracking loop loss rate.

[0114] The fusion decision unit 62 is configured to use a logistic regression model to adaptively fuse each index, and output a detection confidence.

[0115] The trigger control unit 63 is configured to automatically call the electromagnetic interference intelligent identification module when the confidence is greater than 90%, and output a preliminary interference classification.

[0116] In the embodiment, the electromagnetic interference detection module 6 determines the existence of electromagnetic interference based on receiver output parameters in real time. The module takes multi-index fusion decision as the core, monitors key indicators such as sudden decrease of carrier-to-noise ratio, correlation peak distortion and sudden increase of tracking loop loss rate, uses a logistic regression model to adaptively fuse each index, automatically calls the electromagnetic interference intelligent identification module when the detection confidence is greater than 90%, and simultaneously outputs a preliminary interference type classification to speed up the subsequent identification process.

[0117] Further, the interference feature database management module 7 comprises a hierarchical storage unit 71, a scene matching unit 72, an incremental learning unit 73 and a data stream output unit 74, which are sequentially connected;

[0118] The hierarchical storage unit 71 is configured to store sample features and scene labels according to interference types.

[0119] The scene matching unit 72 is configured to automatically load a matching scene library according to the geographic location of the receiver.

[0120] The incremental learning unit 73 is configured to support administrators to add new samples and update feature indexes through an interface.

[0121] The data stream output unit 74 is configured to provide a training sample set to the identification module in a standardized JSON format.

[0122] In the embodiment, the interference feature database management module 7 stores sample features according to interference types (narrowband, wideband, impulse, sweep frequency), including center frequency, bandwidth, time width, modulation parameters and scene labels (airport / port / city), and automatically loads a matching scene library according to the geographic location when the receiver is started (for example, ship radar interference samples are preferentially called in a port environment). The module supports administrators to add new samples and update feature indexes through an incremental learning interface, and finally provides a training sample set to the identification module in a standardized JSON data stream.

[0123] Further, the electromagnetic interference intelligent identification matching module 8 comprises a training unit 81, a classification identification unit 82, a sample matching unit 83 and a result output unit 84, which are sequentially connected, and the sample matching unit 83 is connected with the result output unit 84.

[0124] The training unit 81 is configured to use sample data in the interference feature database to train a classifier model.

[0125] The classification identification unit 82 is configured to input a feature vector extracted by the time-frequency domain module into the classifier to obtain a probabilistic interference type label.

[0126] The sample matching unit 83 is configured to match the most similar historical sample in the database based on a K-nearest neighbor algorithm.

[0127] The result output unit 84 is configured to output an interference type, a confidence and a matching sample ID to drive an anti-interference strategy generation.

[0128] In the embodiment, the electromagnetic interference intelligent identification matching module 8 is divided into two stages of training and identification. The training stage uses database samples to train the SVM / random forest classifier, and the feature dimensions include the number of spectral peak values, the instantaneous bandwidth, the peak-to-average ratio in the time domain, etc. In the identification stage, the feature vector extracted by the time-frequency domain module is first input into the classifier to obtain a probabilistic type label, and then the most similar historical sample in the database is matched based on the K nearest neighbor (KNN) algorithm, and finally the interference type, confidence and matching sample ID are output to drive the anti-interference strategy generation.

[0129] Please refer to Figure 10 In a second aspect, a GNSS interference anomaly detection and identification method is used in the GNSS interference anomaly detection and identification system of the first aspect, and includes the following steps:

[0130] S1 receives and solves GNSS intermediate frequency signals to output basic observation parameters;

[0131] Specifically, the Doppler frequency offset and the initial code phase of the signal are quickly estimated, the histogram statistical method is used to lock the bit transition boundary of the navigation message, the phase-locked loop (PLL) and the delay-locked loop (DLL) are combined to dynamically track the carrier and code phase, the carrier-to-noise ratio (C / N0), the Doppler frequency shift, the carrier phase, and the pseudo-range are output in real time, the ephemeris and clock correction data in the navigation message are analyzed, the least squares method or Kalman filtering is used to fuse the pseudo-ranges of multiple satellites, and the position, velocity, and time (PVT) results are output.

[0132] S2 detects the influence of ionospheric scintillation and solar radio burst on the signal;

[0133] Specifically, the carrier-to-noise ratio, the carrier phase error, and the phase-locked loop jitter parameters output by the receiver are analyzed, the amplitude scintillation index and the phase scintillation index are extracted, the threshold is dynamically adjusted in combination with the geographic latitude and the Kp index, the scintillation intensity level, the occurrence time, and the duration are output, it is monitored whether the multi-frequency point carrier-to-noise ratio appears synchronous, non-locality sudden drop (>10 dB-Hz), the interference time is matched with the solar flare burst time, the artificial interference is excluded through the similarity of multi-frequency point interference, the affected satellite is marked and the positioning result credibility is reduced.

[0134] S3 identifies satellite anomalies;

[0135] Specifically, the difference between the pseudo-range and the carrier phase change rate is monitored, the correlation peak asymmetry is detected by using a multi-correlator, the navigation message CRC and the ephemeris parameter rationality are checked, the multi-satellite consistency test is performed by using the geometric distance residual analysis, the abnormal satellite PRN number and type are output, and the receiver is guided to exclude the faulty satellite.

[0136] S4 multi-index fusion judges the existence of electromagnetic interference;

[0137] Specifically, the key indicators such as sudden increase of carrier-to-noise ratio, correlation peak distortion, tracking loop lock rate, etc. are monitored, a logistic regression model is used to adaptively fuse multiple indicators, a detection confidence is generated, when the confidence is greater than 90%, an electromagnetic interference intelligent identification module is automatically called, and a preliminary interference type classification is output.

[0138] S5 time domain / frequency domain / time-frequency joint analysis of interference characteristics;

[0139] Specifically, the amplitude histogram is calculated to identify impulse interference, the autocorrelation function is used to detect periodic interference, the FFT spectrum analysis is used to locate narrowband / single-frequency interference, the spectral entropy value is used to reduce energy concentrated interference, the short-time Fourier transform (STFT) is used to determine the interference period and frequency band, the wavelet transform is used to capture the transient Chirp signal characteristics, and the bandwidth, center frequency, modulation type and other parameters of the interference are output.

[0140] S6 machine learning classifies the interference type and matches the historical sample;

[0141] Specifically, the SVM / random forest classifier is trained using the interference characteristic database sample in the training stage, the feature dimensions include the spectral peak number, the instantaneous bandwidth, and the time domain peak-to-average ratio. In the identification stage, the feature vector extracted by the time-frequency domain module is input into the classifier to obtain a probabilistic type label. The K nearest neighbor (KNN) algorithm is used to match the most similar historical sample in the database. The interference type, confidence and matching sample ID are output to drive the anti-interference strategy generation.

[0142] S7 activate ARAIM algorithm to reduce VPL when satellite geometry deteriorates;

[0143] Specifically, the present application reduces VPL by reducing the upper and lower limits of VPL, and the weighting of the solution containing two maximum faults can also reduce VPL when other conditions are met. In the algorithm, the all-view solution is replaced by the weighted value of a single constellation solution to achieve the reduction of the protection level. In ARAIM, the final VPL value depends on all possibilities in the all-view. When the VPL of a certain period exceeds the threshold or is too large, it indicates that the satellite geometry has been very poor. At this time, the The difference caused by the two worst geometries is improved. Hereinafter, it is referred to as EA-ARAIM.

[0144] Generally, the all-view solution should satisfy the following criteria:

[0145] (Formula 1)

[0146] Here, EA-ARAIM is to try to optimize the solution of VPL by an idea of extremum approximation, and the solution is as follows:

[0147] (Form 2)

[0148] wherein, and is an uncertain parameter less than 1, and are two projection matrices that differ most from the full-view position solution.

[0149] After re-computing the parameters, the following objective function is obtained:

[0150] (Form 3)

[0151] The above parameters are nominally replaced to facilitate subsequent analysis and computation, such as using in the above expressions. The new definitions are:

[0152] (Form 4)

[0153] The problem is then transformed to:

[0154] (Form 5)

[0155] In fact, it is very complicated to find the optimal solution to Form 5 using traditional methods, which is a high-dimensional linear equation. Note that it is completely possible to satisfy Form 1 for the full-view solution. Then it is a feasible idea to find an approximation to replace the old full-view solution. Combining Form 1, we can get:

[0156] (Form 6)

[0157] Substitute the new full-view solution into the solution and rewrite Form 3 to get:

[0158] (Form 7)

[0159] It can be found that the amount of calculation can be greatly simplified because and have a strong correlation. The following work is to find a suitable t value that satisfies the constraint limit while making the VPL value as small as possible. That is:

[0160] (Form 8)

[0161] Through the above formulas, it can be seen that the solution to Form 8 will become very simple at this time, because the complex multi-dimensional problem is transformed into a simple one-dimensional problem, which can be easily solved in a numerical way. This algorithm reduces the upper and lower limits of VPL. Naturally, it will reduce the final determined VPL. During the algorithm process, when the full-view solution is changed, the The VPLs will also change. In fact, the algorithm takes an approximation vector to solve the equation. Therefore, the optimal allocation of the actual But the final result of the algorithm is still satisfactory. The ARAIM algorithm is run at each epoch, and the EA-ARAIM algorithm is only started when the ARAIM algorithm is not sufficient to meet the PLV-200 performance. As can be seen from equation 6, the EA-ARAIM algorithm leads to a redistribution of the For certain failure modes, this redistribution will undoubtedly lead to a smaller According to the satellite exclusion principle, it is possible to meet the condition of excluding the satellite representing this failure mode, even if this does not occur in the test. If this occurs in the constellation failure mode, it will lead to a terrible situation, i.e. a certain constellation will not participate in positioning, thus leading to a large deviation in the positioning result. Therefore, when the basic algorithm of the ARAIM user is available, the improved algorithm is not used, and when EA-ARAIM is used, if the satellite failure occurs again, even if the probability is very small, the algorithm needs to be abandoned. However, this will not affect the overall performance of the EA-ARAIM, because this situation is a completely small probability event relative to the ARAIM during the entire operation period. It is worth noting that this possibility is not mentioned in other improved ARAIM algorithms.

[0162] The present application also provides an improved ARAIM algorithm based on a feedback structure, which optimizes the protection level through a feedback idea and obtains the VPL value through a new position estimator.

[0163] The new position estimator can be a linear combination of the existing projection matrices. Therefore, the new all-view solution can be expressed as:

[0164]

[0165] wherein, is a linear combination of the projection matrices under the GPS and GAL constellation failures without restrictions.

[0166] When does not participate in the calculation of the new position estimator. The above equation becomes a linear combination of the projection matrices between the two constellation failures. It is very convenient to calculate the subsequent parameters using the new position estimator under such a combination, and the amount of calculation is very small. Because in this case, a one-dimensional equation is obtained. Therefore, the weight between the constellation failures when solving the optimal VPL is completely uncomplicated.

[0167] However, in general, the new position estimator needs contribution. Because the accuracy under the constellation failure is the worst. There is Even if the best VPL weight is obtained, the weight obtained is usually abandoned because the accuracy requirement is not met. Its purpose is to ensure the required accuracy while sacrificing very little VPL, thereby increasing the usability of ARAIM. From the extreme value-realistic ARAIM algorithm, we can see that... The involvement of [the system / mechanism] will complicate all subsequent calculations. Furthermore, the symmetric relationship between constellation fault weights will be broken. In other words, the final weights between the GPS and GAL constellations will no longer be a simple linear relationship. Therefore, to address this problem, this invention proposes a search method based on a feedback structure.

[0168] Simplifying the previous formula, we get:

[0169]

[0170] along with Towards As they gradually approach each other, the accuracy will gradually decrease. New The new accuracy requirement can be obtained as follows:

[0171]

[0172] The above expression, after being rearranged, is similar to:

[0173]

[0174] at this time The parameters of a quadratic polynomial are redefined as:

[0175]

[0176] Solving the equation yields:

[0177]

[0178] It is not difficult to find through analysis that and (If c is positive, optimization cannot continue because the precision requirement cannot be met.) We can see that the lower bound is negative and the upper bound is positive. Also, for the equation, the coefficients should be as close as possible. The final result will be more ideal, which means hope. Get as close to 1 as possible. There are two possibilities: the upper limit is greater than 1 or less than 1. If it's higher than that value, then... It is feasible and can be selected; otherwise, if it is below that value, the upper limit is selected. Therefore, we can obtain:

[0179]

[0180] The calculation method after determining the constellation fault weighting parameter is analyzed above, and the feedback-based weighting parameter determination method is introduced below.

[0181] ①First, assume that the final weighting parameter is 0, that is:

[0182]

[0183] ②The distribution threshold under the current "optimal" can be obtained by calculation , and the new weight will be obtained by distributing the threshold value .

[0184] ③Multiply and the weight under the "optimal" distribution threshold to obtain new and , and then recalculate the distribution threshold and record it as . At this time, compare and . If the deviation is greater than , continue to repeat steps ① to ③. Otherwise, it means that the corresponding optimal weight has been found at this time.

[0185] S8 outputs the suppression scheme by integrating all detection results.

[0186] Specifically, the system integrates the electromagnetic interference type, confidence, and matching sample ID output by each module. Satellite anomaly type and PRN number (from the satellite signal anomaly identification module 4). Ionosphere / solar radio influence evaluation. Optimized VPL value (from EA-ARAIM / feedback ARAIM). According to the interference type, call the historical anti-interference strategy (such as time domain filtering for pulse interference, and enable frequency domain notch for narrowband interference). Exclude the faulty satellite combined with the satellite anomaly information, and use the ARAIM optimization result to ensure the integrity of positioning.

[0187] The above only discloses a preferred embodiment of the GNSS interference anomaly detection and identification system and method of the present application, and of course cannot limit the scope of the rights of the present application. Those skilled in the art can understand that the above-mentioned embodiment can be implemented in whole or in part, and equivalent changes made according to the claims of the present application still fall within the scope of the present application.

Claims

1. A GNSS interference anomaly detection identification system, characterized in that, The GNSS interference anomaly detection and identification system comprises a multi-system multi-frequency point receiver module, an ionospheric scintillation identification module, a solar radio identification module, a satellite signal anomaly identification module, a time-frequency domain analysis module, an electromagnetic interference detection module, an interference feature database management module and an electromagnetic interference intelligent identification matching module, the ionospheric scintillation identification module, the solar radio identification module, the satellite signal anomaly identification module, the time-frequency domain analysis module and the electromagnetic interference detection module are connected with the multi-system multi-frequency point receiver module, the electromagnetic interference intelligent identification matching module is connected with the interference feature database management module and the electromagnetic interference detection module; The multi-system multi-frequency point receiver module is used for receiving GNSS intermediate frequency signals, performing acquisition, bit synchronization, tracking and message decoding, and outputting carrier-to-noise ratio, Doppler, carrier phase, pseudo-range and positioning results; The ionospheric scintillation identification module is used for identifying whether the GNSS signals are affected by ionospheric scintillation according to the signal processing parameters output by the multi-system multi-frequency point receiver module, and giving an estimation of ionospheric scintillation characteristic quantity; The solar radio identification module is used for identifying whether the GNSS signals are affected by solar radio burst according to the signal processing parameters output by the multi-system multi-frequency point receiver module; The satellite signal anomaly identification module is used for judging whether the satellite signal itself is abnormal according to the signal processing parameters output by the multi-system multi-frequency point receiver module; The time-frequency domain analysis module is used for performing time-frequency domain analysis on the received GNSS intermediate frequency signals, and extracting time-frequency domain characteristic parameters of the electromagnetic interference signals; The electromagnetic interference detection module is used for judging whether there is electromagnetic interference in the signals according to the signal processing parameters output by the multi-system multi-frequency point receiver module, and calling the electromagnetic interference intelligent matching identification module according to the detection result; The interference feature database management module is used for storing characteristic sample data of different types of interference signals in multiple scenes, and transmitting interference feature training data to the electromagnetic interference identification module according to the scene of the received intermediate frequency signals; The electromagnetic interference intelligent identification matching module is used for training and learning data samples of different types of electromagnetic interference in a given scene, and classifying the to-be-tested samples composed of characteristic parameters according to a training function.

2. The GNSS interference anomaly detection and identification system according to claim 1, wherein The multi-system multi-frequency point receiver module comprises a signal acquisition unit, a bit synchronization unit, a signal tracking unit and a positioning calculation unit, and the signal acquisition unit, the bit synchronization unit, the signal tracking unit and the positioning calculation unit are connected in sequence; The signal acquisition unit is used for quickly estimating Doppler frequency offset and initial code phase, and completing signal acquisition; The bit synchronization unit is used for locking the bit transition boundary of navigation message through histogram statistics method; The signal tracking unit is used for jointly tracking carrier and code phase through phase-locked loop and delay-locked loop, and outputting carrier-to-noise ratio, Doppler shift, carrier phase and pseudo-range in real time; The positioning solution unit is configured to analyze ephemeris and clock correction data in a navigation message, fuse pseudo ranges of multiple satellites by using a least square method or Kalman filtering, and output a position, a speed, and a time result. 3.The GNSS interference anomaly detection and identification system of claim 1, wherein The ionospheric scintillation identification module comprises a parameter analysis unit, a feature extraction unit, and a dynamic decision unit, and the parameter analysis unit, the feature extraction unit, and the dynamic decision unit are sequentially connected. The parameter analysis unit is configured to analyze carrier-to-noise ratio, carrier phase error, and phase-locked loop jitter parameters output by a receiver. The feature extraction unit is configured to calculate an amplitude scintillation index and a phase scintillation index. The dynamic decision unit is configured to dynamically adjust a threshold value in combination with a geographic latitude and a Kp index, and output a scintillation intensity level, a time of occurrence, and a duration. 4.The GNSS interference anomaly detection and identification system of claim 1, wherein The solar radio identification module comprises a carrier-to-noise ratio monitoring unit, a time matching unit, and an interference confirmation unit, and the carrier-to-noise ratio monitoring unit, the time matching unit, and the interference confirmation unit are sequentially connected. The carrier-to-noise ratio monitoring unit is configured to detect whether multi-frequency carrier-to-noise ratios appear synchronously and non-locally. The time matching unit is configured to match an interference occurrence time with a solar flare burst time. The interference confirmation unit is configured to exclude local human interference by multi-frequency point interference similarity, mark affected satellites, and reduce the weight of positioning results. 5.The GNSS interference anomaly detection and identification system of claim 1, wherein The satellite signal anomaly identification module comprises a code-carrier separation detection unit, a signal distortion judgment unit, a data verification unit, and a fault positioning unit, and the code-carrier separation detection unit, the signal distortion judgment unit, and the data verification unit are respectively connected with the fault positioning unit. The code-carrier separation detection unit is configured to monitor a pseudo range and a carrier phase rate difference. The signal distortion judgment unit is configured to detect correlation peak asymmetry by using a multi-correlator. The data verification unit is configured to verify navigation message CRC and ephemeris parameter rationality. The fault positioning unit is configured to position a single satellite fault by using a geometric distance residual analysis, and output an abnormal satellite PRN number and type. 6.The GNSS interference anomaly detection and identification system of claim 1, wherein The time-frequency domain analysis module comprises a time domain analysis unit, a frequency domain analysis unit, a time-frequency joint analysis unit, and a parameter output unit, and the time domain analysis unit, the frequency domain analysis unit, and the time-frequency joint analysis unit are respectively connected with the parameter output unit. The time domain analysis unit is configured to identify pulse interference by calculating an amplitude histogram, and detect periodic interference by using an autocorrelation function. The frequency domain analysis unit is configured to position narrowband / single frequency interference by using FFT spectrum analysis, and reduce energy concentration type interference by using a spectrum entropy value. The time-frequency joint analysis unit is configured to determine an interference time period and a frequency band by using a short-time Fourier transform, and capture transient Chirp signal characteristics by using a wavelet transform. The parameter output unit is configured to output an interference bandwidth, a center frequency, and a modulation type parameter. 7.The GNSS interference anomaly detection and identification system of claim 1, wherein, The electromagnetic interference detection module comprises an index monitoring unit, a fusion decision unit, and a trigger control unit, which are connected in sequence. The index monitoring unit is configured to monitor a carrier-to-noise ratio (C / N) drop, a correlation peak distortion, and a tracking loop loss rate increase in real time. The fusion decision unit is configured to use a logistic regression model to adaptively fuse the indexes and output a detection confidence. The trigger control unit is configured to automatically call an electromagnetic interference intelligent identification module when the confidence is greater than 90% and output a preliminary interference classification. 8.The GNSS interference anomaly detection and identification system of claim 1, wherein, The interference feature database management module comprises a hierarchical storage unit, a scene matching unit, an incremental learning unit, and a data stream output unit, which are connected in sequence. The hierarchical storage unit is configured to store sample features and scene labels according to interference types. The scene matching unit is configured to automatically load a matching scene library according to a receiver geographic location. The incremental learning unit is configured to support an administrator to add new samples and update feature indexes through an interface. The data stream output unit is configured to provide a training sample set to the identification module in a standardized JSON format. 9.The GNSS interference anomaly detection and identification system of claim 1, wherein, The electromagnetic interference intelligent identification matching module comprises a training unit, a classification identification unit, a sample matching unit, and a result output unit, which are connected in sequence, and the sample matching unit is connected to the result output unit. The training unit is configured to use sample data in the interference feature database to train a classifier model. The classification identification unit is configured to input a feature vector extracted by the time-frequency domain module into the classifier to obtain a probabilistic interference type label. The sample matching unit is configured to match the most similar historical sample in the database based on a K-nearest neighbor algorithm. The result output unit is configured to output an interference type, a confidence, and a matching sample ID to drive an anti-interference strategy generation.

10. A GNSS interference anomaly detection and identification method for the GNSS interference anomaly detection and identification system of any one of claims 1-9, characterized in that, The method comprises the following steps: Receiving and resolving GNSS intermediate frequency signals to output basic observation parameters; Detecting the influence of ionospheric scintillation and solar radio burst on signals; Identifying satellite abnormalities; Multi-index fusion to determine the existence of electromagnetic interference; Time domain / frequency domain / time-frequency joint analysis of interference characteristics; Machine learning to classify interference types and match historical samples; Activating ARAIM algorithm to reduce VPL when satellite geometry deteriorates; Outputting a suppression scheme based on all detection results.

Images