Edge Computing Based GPS Spoofing Detection Methods

Qian Wang,Zhaojun Lu,Mingze Gao,Gang Qu

DOI: https://doi.org/10.1109/icdsp.2018.8631600

2018-01-01

Abstract:Civilian GPS signals are becoming essential for the navigation systems in many applicators such as vehicles and smart phones. However, from a security perspective, GPS signals are vulnerable to the spoofing attacks. Several spoofing detection methods have been proposed, but most of them require extensive signal processing capabilities and additional equipment such as receivers. These add-ons may not be available for vehicles and smart phones. In this paper, we propose a novel edge computing based approach to reconstruct the lost GPS signal. The basic idea is to collect information at the edge nodes and use them to cross-validate the GPS signals received from the satellite. If there is any evidence of spoofing attacks, our method can reconstruct the GPS signal when the signal is unavailable or not trustworthy. Thus, this method could serve as a backup plan to cope with the failure of GPS for the navigation system. Based on real driving data, we can reconstruct the driving routes with an average error of 10 meters and 6 meters by two different methods. This is sufficiently accurate to detect all the simulated GPS spoofing attacks.