Search

CN-121978717-A - GNSS suppression interference number identification method, device, equipment and medium

CN121978717ACN 121978717 ACN121978717 ACN 121978717ACN-121978717-A

Abstract

The application relates to a GNSS suppression interference number identification method, device, equipment and medium, and relates to the technical field of satellite navigation interference resistance. The method comprises the steps of receiving airspace signals through an array antenna, obtaining array baseband signals through down-conversion and analog-to-digital conversion, constructing an airspace correlation matrix in a complex matrix form based on signal vectors and conjugate transposition operation at a plurality of snapshot moments, extracting all elements of a triangle part on the matrix, arranging the elements according to a preset sequence to form complex vectors as input data, outputting classification results by utilizing a pre-trained deep learning model, and determining the suppression interference number. The method realizes data simplification and complete information reservation by means of the characteristics of the complex matrix, adapts to model input, does not need subjective threshold through deep learning, solves the limitation of the traditional method, has high robustness and accuracy under the environment of low signal-to-interference-and-noise ratio and small snapshot number, performs offline training and real-time reasoning, reduces the calculation complexity, meets the real-time requirement of a terminal, and provides reliable support for an anti-interference algorithm.

Inventors

  • SUN GUOPING
  • OU HONGBO

Assignees

  • 湖南博尚电子科技有限公司

Dates

Publication Date
20260505
Application Date
20260123

Claims (10)

  1. 1. A method for identifying a number of GNSS hold-down disturbances, the method comprising: Acquiring a radio frequency signal, wherein the radio frequency signal is obtained by receiving a space domain signal by a GNSS array antenna, and sequentially performing down-conversion and analog-to-digital conversion on the radio frequency signal to obtain an array baseband signal; According to the array baseband signals, calculating a space domain correlation matrix, wherein the space domain correlation matrix is a Hermite complex matrix constructed by multiplying array received signal vectors at a plurality of snapshot moments with conjugate transposed vectors of the array received signal vectors; Converting the airspace correlation matrix into a format suitable for deep learning model input through preprocessing to generate input data, wherein preprocessing the airspace correlation matrix comprises the steps of extracting all elements of an upper triangle part or a lower triangle part of the airspace correlation matrix, and arranging all the extracted elements according to a preset sequence to form complex vectors to obtain the input data; And inputting the input data into a pre-trained deep learning model to obtain a classification result corresponding to the interference number, and determining the suppression interference number existing in the current environment according to the classification result.
  2. 2. The method for identifying the number of GNSS hold-down interference according to claim 1, wherein the GNSS array antenna is an M-array element uniform linear array, where M is a positive integer greater than 1; the analog-to-digital conversion is ADC sampling, and the array baseband signal obtained after sampling is a digital baseband signal.
  3. 3. The GNSS suppression interference number identification method of claim 1, wherein the calculation formula of the spatial correlation matrix is expressed as: in the above-mentioned description of the invention, Representing the spatial correlation matrix in question, Represent the first The array of snapshot times receives a signal vector, In order to take the number of shots in a short time, Representation pair And performing conjugate transposition operation.
  4. 4. The method for identifying the number of GNSS compacting interferences according to claim 1, wherein the extracting all elements of the upper triangle part or the lower triangle part of the spatial correlation matrix, arranging all the extracted elements in a preset order to form complex vectors, and obtaining the input data includes: extracting all elements in a triangle part on a main diagonal of the airspace correlation matrix; And arranging all the extracted elements according to a preset sequence of line priority to form complex vectors.
  5. 5. The GNSS suppression interference number identification method of claim 1, wherein the deep learning model is one of a convolutional neural network, a fully-connected neural network, a cyclic neural network, or a Transformer architecture.
  6. 6. The method of claim 5, wherein when the deep learning model is a fully connected neural network, the fully connected neural network comprises an input layer, at least one hidden layer and an output layer, the hidden layer uses a ReLU activation function, the output layer uses a Softmax activation function, and the number of neurons of the output layer is , wherein, A preset maximum number of disturbances is indicated and a scenario of 0 disturbances is included.
  7. 7. The GNSS hold-down interference number identification method of claim 5 wherein training the deep learning model comprises: Generating array receiving signal samples under a plurality of different interference scenes through simulation or actual measurement, wherein each signal sample is marked with a real interference number label, and the interference scenes comprise combinations of different interference numbers, interference incoming wave directions, a signal to interference ratio (JSR) and GNSS useful signal incoming wave directions; Respectively calculating and preprocessing the spatial correlation matrix for each signal sample to obtain corresponding training data; and taking the training data as input of the deep learning model and the corresponding interference number labels as expected output, and training the deep learning model in a supervised training mode until the model converges to obtain a pre-trained deep learning model.
  8. 8. A GNSS hold-down interference number identification device, the device comprising: The system comprises a radio frequency signal receiving and processing module, a receiving module and a receiving module, wherein the radio frequency signal receiving and processing module is used for acquiring a radio frequency signal, the radio frequency signal is obtained by receiving a space domain signal by a GNSS array antenna, and the radio frequency signal is subjected to down-conversion and analog-to-digital conversion in sequence to obtain an array baseband signal; The airspace correlation matrix calculation module is used for calculating airspace correlation matrix according to the array baseband signals, wherein the airspace correlation matrix is Hermite complex matrix constructed by multiplying array received signal vectors based on a plurality of snapshot moments with conjugate transpose vectors of the array received signal vectors; The preprocessing module is used for converting the airspace correlation matrix into a format suitable for deep learning model input through preprocessing to generate input data, wherein the preprocessing of the airspace correlation matrix comprises the steps of extracting all elements of an upper triangle part or a lower triangle part of the airspace correlation matrix, and arranging all the extracted elements according to a preset sequence to form complex vectors to obtain the input data; and the suppression interference number determining module is used for inputting the input data into a pre-trained deep learning model to obtain a classification result corresponding to the interference number, and determining the suppression interference number existing in the current environment according to the classification result.
  9. 9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 7 when the computer program is executed.
  10. 10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 7.

Description

GNSS suppression interference number identification method, device, equipment and medium Technical Field The application relates to the technical field of satellite navigation anti-interference, in particular to a GNSS suppression interference number identification method, device, equipment and medium. Background GNSS (e.g., GPS, beidou, GLONASS, galileo, etc.) plays a vital role in providing accurate positioning, navigation, and timing services for global users. However, since GNSS signals arrive very weak from space to the ground, they are very susceptible to man-made interference, especially powerful jamming. The squelch interference can overlay or drown out weak GNSS useful signals, causing the receiver to fail to acquire and track satellite signals properly, resulting in a positioning failure. To combat the interference, spatial domain interference rejection techniques based on array antennas have become the dominant solution. The technology forms null in the interference incoming wave direction by adjusting the weight vector of the array antenna, thereby inhibiting the interference. However, most advanced immunity algorithms (e.g., feature subspace-based algorithms) require knowledge of the exact number of interferers in advance. If the number of interferences is estimated incorrectly, the interference suppression performance may be seriously degraded, and even the useful signal may be suppressed as interference by mistake. Currently, the interference number estimation method is mainly based on information theory criteria such as Akaike information criteria (Akaike Information Criterion, AIC) and minimum description length criteria (Minimum Description Length, MDL). These methods determine the number of interferences by analyzing eigenvalues of the covariance matrix of the received signal. However, such conventional methods have significant drawbacks of (1) sensitivity to signal-to-interference-and-noise ratio (Signal to Interference plus Noise Ratio, SINR), i.e., blurred eigenvalue distribution in low SINR environments, and a dramatic drop in estimation accuracy of the conventional methods. (2) The threshold is required to be subjectively set, the threshold is required to be set according to experience by part of the improved algorithm, the adaptability is poor, and the performance is unstable under different scenes. (3) The computing complexity is high, complex operations such as eigenvalue decomposition and the like are needed, and real-time performance is difficult to ensure for GNSS terminals with limited computing resources. Disclosure of Invention Accordingly, in order to solve the above-mentioned problems, it is necessary to provide a method, a device, an apparatus and a medium for identifying the number of GNSS compacting interferences, which are adaptable to a complex electromagnetic environment, have high estimation accuracy and have good real-time performance. A GNSS hold-down interference number identification method, the method comprising: Acquiring a radio frequency signal, wherein the radio frequency signal is obtained by receiving a space domain signal by a GNSS array antenna, and sequentially performing down-conversion and analog-to-digital conversion on the radio frequency signal to obtain an array baseband signal; According to the array baseband signals, calculating a space domain correlation matrix, wherein the space domain correlation matrix is a Hermite complex matrix constructed by multiplying array received signal vectors at a plurality of snapshot moments with conjugate transposed vectors of the array received signal vectors; Converting the airspace correlation matrix into a format suitable for deep learning model input through preprocessing to generate input data, wherein preprocessing the airspace correlation matrix comprises the steps of extracting all elements of an upper triangle part or a lower triangle part of the airspace correlation matrix, and arranging all the extracted elements according to a preset sequence to form complex vectors to obtain the input data; And inputting the input data into a pre-trained deep learning model to obtain a classification result corresponding to the interference number, and determining the suppression interference number existing in the current environment according to the classification result. In one embodiment, the GNSS array antenna is an M-array element uniform linear array, where M is a positive integer greater than 1; the analog-to-digital conversion is ADC sampling, and the array baseband signal obtained after sampling is a digital baseband signal. In one embodiment, the calculation formula of the spatial correlation matrix is expressed as: in the above-mentioned description of the invention, Representing the spatial correlation matrix in question,Represent the firstThe array of snapshot times receives a signal vector,In order to take the number of shots in a short time,Representation pairAnd performing conjugate transposition