Cyber-Physical Trust Binding for NextG Networks and Applications

Abstract

Next Generation (NextG) networks are evolving into cyber-physical systems that tightly intertwine the digital and physical worlds, enabling transformative applications such as connected autonomous vehicles (CAVs) and the massive Internet of Things (IoT). However, this convergence introduces a critical {\em cyber-physical trust gap:} traditional cryptographic security, which verifies digital identities and message integrity, cannot ensure that digital information faithfully represents the physical world’s state and behavior. This dissertation addresses this fundamental gap by proposing a unified framework that binds digital identities and communications to verifiable physical properties, establishing trust across both outdoor macro-scale and indoor micro-scale applications. For the CAV domain, we develop mechanisms that ensure transmitted data corresponds to physical reality. First, we introduce Proof-of-Following (PoF), a physical access control protocol that verifies a vehicle’s presence within a platoon by correlating ambient RF signals, thereby preventing remote impersonation attacks. Second, we present Wiggle, a physical challenge–response protocol that employs motion perturbations to verify a vehicle’s relative position and lane, thwarting closely following adversaries. Third, to mitigate data falsification attacks from malicious insiders, we design a trust-aware Partially Observable Markov Decision Process (POMDP) framework that fuses sensor and V2V data while dynamically evaluating message veracity, ensuring safety and efficiency under adversarial conditions. For the IoT domain, we establish cyber-physical trust by embedding physical-layer randomness into security primitives. We first develop a reconfigurable antenna-based integrity protocol that defends against sophisticated signal manipulation attacks, providing message protection without pre-shared keys. Finally, we propose a blockage-driven randomness extraction mechanism for millimeter-wave (mmWave) systems that harvests entropy from random channel blockages, enabling secure key generation resilient to co-located passive eavesdropping and active man-in-the-middle adversaries.