Distributed State Estimation for Hidden Markov Models with Dynamic Quantization and Rate Allocation

Authors:	Huang Minyi, University of Melbourne, Australia Dey Subhrakanti, University of Melbourne, Australia
Topic:	1.4 Stochastic Systems
Session:	Control of Stochastic Systems
Keywords:	Hidden Markov models, state estimation,sensor networks, dynamic quantization.

Abstract

This paper considers state estimation of hidden Markov models bysensor networks. By employing feedback from the fusion center tothe sensor nodes, a dynamic quantization scheme is proposed andanalyzed by a stochastic control approach. Dynamic rateallocation is also considered.