Optimal Jamming Attack Schedule Against Wireless State Estimation in Cyber-Physical Systems

Recently, the security issues of Cyber-Physical Systems (CPS) have gained a growing amount of research attention. As a typical application of CPS, remote state estimation systems over wireless channels are vulnerable to various cyber attacks, such as channel jamming attacks. Standing on the point of the jamming attacker, this paper investigates the problem of optimal attack schedule under energy constraints against a wireless state estimation system, where two sensors transmit data to a remote estimator over two independent wireless channels. Due to its radio constraint, the attacker is assumed to only launch jamming attack on one of the two channels at a time. We formulate the problem of optimal attack schedule as a nonlinear program which aims to maximize the remote estimation error covariance subjecting to the attacker's energy constraint. We then theoretically derive the optimal schedule and shows that it depends on the attacker's energy budget, the physical system dynamics and the channel properties. Finally, the theoretical results are validated and evaluated through numerical simulations.