Optimal Sensor Schedule Design with An Unknown Attack Paradigm

In this paper, we mainly consider a scheduling problem with respect to CPS (Cyper-Physical System), which aims to minimize the estimation error over a finite time horizon under an unknown attack paradigm and a limited energy budget. Our goal of this paper is to decide whether to transmit data packet or not at the current time step. Due to the presence of an attacker, there is a certain probability that the sensor may fail to successfully transmit the packet. In order to solve the problem, we first define a series of sub-schedules and sub-costs to describe the accumulated impact on the global cost at different times, and based on which, a necessary condition of optimality is firstly proposed. Then, we discuss the effect brought by the initial value and present a detailed algorithm for generating the optimal scheduling sequence. Numerical simulations have shown the correctness and effectiveness of our proposed optimal schedule.