A Radar System With Adaptive Waveform Selection Against Dynamic Spoofing Attacks

With the spread of automotive radar systems, spoofing among the frequency-modulated continuous-wave (FMCW) millimeter-wave (mmWave) radars in autonomous vehicles (AVs) is becoming a severe issue. In this work, a dynamic spoofing attack scenario in the real world is constructed based on the fact that spoofing is a minority situation. Since conventional radars miss genuine targets in a spoofing environment and antispoofing radars perform far worse than conventional radars without spoofing, the separate application of both radars in our proposed scenario does not yield satisfactory results. To resolve this issue, we propose a cognitive radar system with adaptive waveform selection against spoofing attacks. First, an authentication mechanism and a quantitative model are constructed to detect malicious attacks, a process to be considered essentially adaptive cognition. Second, the appropriate waveform to be emitted is selected based on the results of the adaptive cognition module. To resist spoofing, we propose a waveform with joint frequency-phase modulation (JFPM) combining frequency hopping and phase coding. Based on the location of the frequency hopping, we categorize it into two schemes: intrachirp JFPM and interchirp JFPM. Finally, simulations and experiments demonstrate that our proposed system can accomplish target detection with relatively high accuracy in our scenario.