A Switching Approach to Packet Loss Compensation Strategy

In networked control systems, a zero or hold compensation strategy is usually used for the missing signal when a packet is lost. In this paper, a switching approach is proposed for the consecutive loss compensation, where the packet loss between sensor and controller is modeled as an i.i.d. Bernoulli process. The held signal was divided into three categories, that is, the current signal, usable signal, and unusable signal, according to the number of consecutive loss. Then, the system is built as a switching system with normal, time-delay, and open-loop subsystems, where the switching rule associates with the packet loss. It is theoretically proved that the proposed switching approach is effective for packet loss compensation. The dividing criteria of the held sensor measurements are derived. The usefulness of the proposed results is verified by a numerical example.