Stabilization for Networked Control Systems with Simultaneous Input Delay and Markovian Packet Losses

TL;DR

This paper addresses the mean square stabilization of discrete-time networked control systems with both input delay and Markovian packet losses, providing new conditions for solvability and stabilization.

Contribution

It introduces a novel equivalence condition for finite-horizon LQ problems and a necessary and sufficient stabilization criterion using coupled algebraic Riccati equations.

Findings

Derived FBSDEs-M for LQ problem solvability

Established CAREs-M for stabilization

Addressed simultaneous delay and packet loss in NCSs

Abstract

The mean square stabilization problem for discrete-time networked control systems (NCSs) is investigated in this article. What the difference from most previous works is that input delay and packet losses occur simultaneously in the communication channel, moreover, the data packet dropout is modeled as a time-homogeneous Markov process which will bring some difficulties in solving the problem due to the temporal correlation. The contributions in this paper can be summarized as two points. Firstly, the equivalence condition for the solvability of linear quadratic optimal problem in finite horizon subject to the discrete-time NCSs is expressed by solving the forward and backward stochastic difference equations (FBSDEs-M) which is derived from the maximum principle involving Markov jump and delay. Secondly, under basic assumption, the necessary and sufficient condition of mean square stabilization is given by the solutions to the coupled algebraic Riccati equations with Markov jump (CAREs-M). To our best knowledge, the problems studied in this paper are new because most previous works mainly discussed the case of only delay or packet dropout in NCSs.