TESTING AND DEFENDING METHODS AGAINST DOS ATTACK IN STATE ESTIMATION

In this paper, we consider a state estimation problem. In this problem, a sensor measures the state of a linear discrete-time system and sends measurements to an estimator via a packet-dropping communication link. We are concerned with the effect of Denial-of-Service (DoS) attacks on stability of the estimation system, and particularly focus on how to examine whether the communication channel is under DoS attack or not as well as how to defend accordingly, if defense is possible. We formulate the detection problem as a hypothesis testing problem provided that the statistics of the communication channel is known a priori. Two defense countermeasures are proposed: one of which uses a secured packet coding approach to partly compensate the previous packet loss; and in the other the sensor's transmission power is raised to resist the jamming effect brought by the DoS attack. Simulations are provided to demonstrate the main results.