Synergistic Cyber–Physical Design for DoS-Resilient Distributed Cooperative Control Under Actuator Saturation

Abstract

This article investigates distributed cooperative tracking of networked vehicles when denial-of-service (DoS) attacks disrupt reference reconstruction and actuator saturation constrains tracking execution. A synergistic cyber-physical framework is developed by integrating a distributed event-triggered estimator (DETE) with a saturation-aware cooperative tracking controller (SCTC). The DETE reconstructs the virtual leader states under intermittent communication, maintains local prediction using stored acceleration information during detected DoS intervals, and guarantees strictly positive inter-event times. Based on the reconstructed References and local vehicle states, the SCTC incorporates the saturation residue, model uncertainties, and external disturbances into a lumped uncertainty, which is estimated and compensated by a super-twisting disturbance observer. Lyapunov analysis establishes uniform ultimate boundedness of the cyber-layer estimation errors and physical-layer tracking errors under graph connectivity and DoS frequency and duration conditions. Simulations and experiments on a six-vehicle platform validate coordinated tracking under intermittent communication and constrained actuation. Note to Practitioners—Networked vehicle formations are increasingly used in inspection, logistics, monitoring, and surveillance, where coordination must be maintained over wireless networks with limited bandwidth and under actuator constraints. Communication denial interrupts the propagation of motion references, while actuator saturation, modeling errors, and external disturbances degrade their physical execution. The proposed framework addresses these coupled effects through two coordinated layers. When a DoS attack is detected, the cyber layer suspends communication updates and predicts the virtual leader states using stored acceleration information; event-triggered transmission is resumed after communication recovery. The physical layer then tracks the reconstructed references using local vehicle measurements, saturation compensation, and disturbance estimation. The event-triggered mechanism reduces unnecessary transmissions while guaranteeing strictly positive inter-event times. Experiments on a six-vehicle indoor platform demonstrate coordinated tracking under intermittent communication and constrained actuation, indicating the implementation potential of the method for vehicle networks subject to detectable packet denial attacks.