Qiang Jia

As intelligent vehicle systems advance, preserving privacy in multi-vehicle coordination is critical to prevent sensitive data exposure. This paper introduces a novel privacy-preserving platooning control framework for multi-vehicle systems, leveraging a directed bipartite graph topology. Unlike encryption or transformation-based methods, each vehicle maintains a virtual replica, broadcasting only replica states and updating via neighbors’ replicas through the bipartite graph’s structure. This design inherently limits information flow, shielding true states (e.g., position, velocity) from internal honest-but-curious neighbors and external eavesdroppers, while avoiding heavy computational overhead. Lyapunov-based analysis ensures successful platooning with exponential convergence, and the control gain is designed via a modified algebraic Riccati equation. Simulations validate performance in 1-D vehicle platoons and 2-D drone formations. With low computational cost and adaptable topology, the method excels in real-time consumer electronics applications, such as autonomous vehicle fleets, coordinated drone swarms, and other IoT-enabled smart systems, offering a lightweight, scalable alternative to conventional privacy-preserving techniques.

This paper presents a novel lightweight privacy-preserving control protocol for leader-following consensus in discrete-time multi-agent systems, tailored for consumer IoT applications such as electric vehicles and unmanned aerial vehicles (UAVs). By employing general decaying signals to mask sensitive user data, including the initial states of agents (e.g. starting positions or velocities), the protocol enables devices to track a leader’s trajectory (e.g., navigation paths for UAV swarms or energy-efficient velocity profiles for vehicular platooning), while safeguarding privacy. By reformulating the problem as the stability problem of a discrete-time system with diminishing disturbances, sufficient tracking conditions are derived based on a new convergence result for infinite sequences. Control gains are efficiently computed via a modified algebraic Riccati equation. Unlike conventional methods relying on computationally intensive encryption or with restrictive exponential decay, the new approach leverages flexible and locally selectable decaying functions, enhancing privacy and adaptability while maintaining a lightweight framework. Numerical simulations of UAV formation and vehicular platooning demonstrate rapid convergence to the leader’s trajectory with lightweight privacy protection robustly against curious agents and eavesdroppers, showcasing the protocol’s superior balance of control accuracy, data security, and computational efficiency, desirable for consumer electronics.