Study on GNSS Spoofing Interference Detection Method in Urban Multipath Environment Based on CNN and Clustering Model

Global Navigation Satellite Systems (GNSS) provide essential positioning, navigation, and time (PNT) information for a wide range of civil and military applications across various scenarios. However, GNSS signals are vulnerable to different forms of interference, particularly spoofing interference. In multipath environments, detecting spoofing interference becomes even more challenging and inefficient. In response to the challenges related to the degradation of spoofing interference detection performance in multipath environments, we propose a novel algorithm named GNSS-CC, which combines the Global Navigation Satellite System (GNSS), Convolutional Neural Network (CNN), and Clustering for the purpose of detecting both multipath interference and spoofing interference. The method leverages a combination of CNN and Clustering technology to construct a model, effectively utilizing distinguishing features between the correlation peaks generated by multipath interference and spoofing interference, enabling accurate classification and identification of multipath and spoofing signals. In this study, various machine learning algorithms are compared, and the simulation results show that GNSS-CC exhibits superior performance, achieving a detection rate of 95.8% for GNSS spoofing interference detection in multipath environments.