Preventing Fake GPS Signal Assaults on Unmanned Vehicles through Contextual Learning - A Blueprint

Civilian-based Global Positioning System (GPS) deployment is insecure, and it can be easily spoofed to report incorrect locations, which may lead threat actors to launch a myriad of attack payloads. In this contribution, we aim at preventing GPS spoofing attacks on unmanned vehicles (UVs) by utilizing machine-learning prototypes that can incorporate surrounding sensory data into the detection and decision-making process. Specifically, we propose to utilize machine-learning models that would feed on a UVf’s immediate scenery text information in order to validate GPS coordinates and thwart any imminent attacks. The machine-learning models rely on and work with proposed geospatial data extracted from UVs in real time. Through experimental simulation results, we observe that our proposed methods are a feasible mitigation approach for preventing GPS spoofing attacks by UVs based on machine learning and artificial intelligence (AI) approaches.