Static Decentralized Control of a Single-Integrator Network with Markovian
Sensing Topology, Sandip Roy, Ali Saberi · 2005
Static stabilization of a decentralized discrete- time single-integrator network that is subject to Markovian variation in the communication/sensing topology is consid- ered. In particular, we develop conditions on the Markovian topology that are sufficient for mean-square stabilization using a decentralized static controller that has no knowledge of the underlying Markov state. Our analysis exposes a deep connection between decentralized control of single-integrator networks with Markovian topology and those with fixed topology: static stabilization of the network with Markovian topology is possible whenever the steady-state time-average of the Markovian topology is amenable to static (fixed-topology- based) decentralized control. I. INTRODUCTION Groups of coordinating autonomous agents are of- ten subject to stochastic fluctuations in their sens- ing/communication capabilities, that may significantly im- pact their ability to complete required tasks. Our aim in this article is to explore static decentralized control of a network of agents with single-integrator dynamics that is subject to Markovian fluctuations in its sensing topology.