Hybrid cyber-resilient control for networked autonomous vehicles

İngilizce (OpenAlex)

We present a novel hybrid control framework for networked autonomous vehicles (AVs) that seamlessly integrates trust-based authentication with a switching-based controller to combat cyber-attacks. The approach combines a cooperative (networked) mode for high performance with a resilient (local) mode that relies solely on onboard sensing when the trust in external data drops below a threshold. We prove, via Lyapunov-based analysis, that each control mode achieves asymptotic stability and that the overall switched system remains stable under typical dwell-time constraints. To validate our design, we simulate a five-vehicle platoon subjected to false data injection (FDI), replay, and denial-of-service (DoS) attacks. The results demonstrate that our method preserves safe inter-vehicle distances and coordinated velocity tracking while recovering quickly from malicious disturbances. Compared to traditional robust or purely cryptographic solutions, our hybrid strategy provides both improved energy efficiency during normal operation and enhanced safety under adversarial conditions.