Optimal power allocation for jamming attack against remote estimation in cyber-physical systems

Abstract

Recently, the security issues of CPS have attracted much attention from many researchers. Standing on the point of attacker, this work investigates the optimal channel jamming attack against wireless state estimation in CPS with energy constraint. Specifically, we consider that the attacker can jam the wireless channel with various power levels that will cause different sensory packet drop probabilities. We formulate the optimal attacker power allocation problem as to maximize the estimation error covariance under the attacker's power constraint. We theoretically derive the optimal power allocation for the attacker, which is shown resembling a parabola shape along time. Finally, we demonstrate the performance of the proposal allocation scheme through simulations.