A GNSS spoofing interference detection method based on convolutional neural network

Abstract

The invention discloses a method for detecting GNSS spoofing interference based on a convolutional neural network, which comprises the following steps: 1) In the signal capture stage, the number N of correlation peaks in the two-dimensional matrix A generated when the signal is captured is larger than the capture threshold peak , if N peak ≥2, it is considered that there is a deception signal, if N peak <2 then proceed to step 2); 2) Intercept the data in the ±2 chip area on the phase axis of the two-dimensional matrix A correlation peak pseudo-code to obtain the detection matrix A s , after data preprocessing, it is sent to the convolutional neural network for detection training and classification, and finally obtains the detection results. The detection method of the present invention has good detection effect and strong applicability, the detection timing is in the signal capture stage, and the complexity is moderate; it can solve the problem that it is difficult to detect when the pseudo-code phase difference ΔT between the fake signal and the real signal is within 2 chips.

Description

technical field [0001] The invention belongs to the technical field of interference detection in satellite navigation systems, in particular to a GNSS deception interference detection method based on a convolutional neural network (CNN). Background technique [0002] Global Navigation Satellite System (GNSS) is a wide-coverage, all-weather, real-time and high-precision navigation system. With the continuous development of satellite navigation technology, GNSS is widely used in various military and civilian facilities. With the development of satellite navigation and positioning technology, the number of users and application scenarios continue to increase, and people pay more and more attention to safety and reliability. The security threats faced by the current satellite navigation system can be mainly divided into unintentional interference and intentional interference. Intentional interference mainly refers to artificial malicious interference, and can be divided into s...

AI Technical Summary

Problems solved by technology

The second is based on signal processing, through signal detection in time domain, frequency domain, power and other characteristics, including signal absolute power detection, carrier-to-noise ratio detection, signal quality monitoring, detection based on RAIM (Receiver Autonomous Integrity Monitoring), etc. This method has strong applicability, but the accuracy of signal absolute power detection and carrier-to-noise ratio detection is poor. RAIM-based detection needs to solve the signal, which has high complexity and low timeliness.

Signal quality monitoring mainly judges whether there is a spoofing signal by whether the signal correlation peak is distorted. The existing signal quality monitoring methods mainly include detecting the number of correlation peaks during signal capture and analyzing the correlator output during signal tracking; the current signal The method of detecting the number of correlation peaks at the time of acquisition is difficult to distinguish the spoofed signal from the real signal when the pseudocode phase difference is less than 2 chips

Images