GNSS Interference Classification using Multi-Layer Perceptron Neural Network Trained by AGPSO

Spoofing attacks are a fundamental threat to civilian Global Navigation Satellite System (GNSS) applications due to their powerful impact on receivers. As a result, plenty of anti-spoofing methods have been developed by researchers in recent years. In this paper, we have proposed a technique for the detection and classification of GNSS interferences based on the so-called Power- Distortion (PD) detector. The PD detector uses received signal power and correlation-profile distortion monitoring for detecting any type of interference including spoofing attacks, jamming, or multi-path. We illustrate that detection and classification can significantly be improved by replacing prior methods of classification with our proposed method. The method uses a Multi-LayerPerceptron Neural Network (MLP NN) trained by Particle Swarm Optimizer with Autonomous Group (AGPSO) in which we will call it MLP NN-AGPSO classifier. The primary usage of this observation is the diagnosis of spoofing attacks among other interferences. The results show that the MLP NN-AGPSO detector exhibits improved detection and classification accuracy. Results obtained from simulation show that AGPSO۳ has better classification performance in comparison to AGPSO۱ and AGPSO۲. More specifically, multi-path signal detection has a great accuracy of ۹۵.۳۳%, and spoofing and jamming are ۹۰.۱۱% ۹۸.۶۷% accurate, respectively.