DDoS attack on wireless cyber–physical systems: A Stackelberg game approach

Abstract

This paper delves into the network security issues of cyber–physical systems (CPSs), where malicious agents can launch distributed denial-of service (DDoS) attacks to disrupt communication channels between sensors and remote estimator. Addressing the strategic interaction between defenders and DDoS attacker, we establish a two-player zero-sum stochastic game framework that portrays the interaction as a dynamic Stackelberg non-cooperative game to deal with the impact of DDoS attacks on CPSs. Using refined Bayesian Nash equilibrium, we determine optimal attack-defense strategies and analyze scenarios with incomplete information. Notably, attacker can access sensor transmission power before acting, prompting us to propose solutions and optimal strategy algorithms. In simulations, we explore three transmission power ranges and thoroughly examine their impact on the utility functions of both attacker and defenders.