Game-Theoretic Transmission Scheduling in CPSs Subjected to DoS Attacks: An Incomplete Information Approach

Abstract

This paper investigates secure transmission scheduling under jamming attacks for remote state estimation in cyber-physical systems (CPSs). The sensor, equipped with a standard Kalman filter, selects a wireless channel at each time step for data transmission according to a prescribed stochastic switching rule, and a denial-of-service (DoS) attacker independently follows its own stochastic rule to choose a channel to jam. Because each agent’s channel selection remains private, the conflict is modeled as a stochastic game under asymmetric information. The analysis distinguishes between two information structures—open-loop and closed-loop—and obtains the corresponding Bayesian Nash equilibrium strategies for both the transmitter and the DoS attacker. Theoretical results characterize how the randomness in channel access, the asymmetry of information, and resource constraints influence the equilibrium behaviors. Numerical simulations validate the analytical findings and illustrate the impact of information availability on estimation accuracy and jamming resilience in CPSs.