Detection of GPS Spoofing Attacks Based on Isolation Forest
Shenzheng Zuo, Yinan Liu, Dongmei Zhang, Pengpeng Xin, Tianxin Liu · 2021 IEEE 9th International Conference on Information, Communication and Networks (ICICN) · 2021
As Global Positioning System(GPS) use non-encryption signals,GPS receivers are vulnerable to spoofing attacks,most of applications of defense base on detect signal power or spatial position,which need expensive hardware.In the recent years,machine learning algorithm have been used to detect the spoofing as a low-cost method. This paper proposes a detection method based on isolation forest algorithm to detect the GPS spoofing.We use observation files from satellite as data source.We use the detection score acquired by the polymerization degree of each data to detect data,which may be normal or abnormal,to determine if the GPS receiver is suffering the spoofing attack.The simulation compares different combinations of input features derived from satellite signals.The results show that the most efficient combination can reach at least 95% accuracy.