Synchronization of neutrally stable linear systems over digital networks

Abstract

This paper studies the synchronization control of linear multi-agent systems with unmeasurable states, finite communication data rate and switching topologies. A class of adaptive quantized-observer based dynamic encoding-decoding schemes for information communication and a class of Certainty Equivalence principle based control protocols for synchronization are proposed for overcoming the uncertainties of network environments. It is shown that if the network topology flow is jointly connected, the communication channels are periodically active, and the agent dynamics are observable and with orthogonal system matrices, then the proposed protocols can drive the dynamic network to synchronization exponentially fast with finite bits of information exchange per step.