Networked state estimation of MIMO Systems

In this paper, the problem of state estimation over multiple parallel communication channels with multiplicative noise is investigated. The main novelty of this work lies in the introduction of the channel/estimator co-design framework which provides the estimator designer an additional freedom to allocate the channel capacities among different output channels. Under this co-design framework, we first study the state observation problem by reaching a conclusion that the minimum total channel capacity required for the output channels such that the estimation error remains bounded is given in terms of the topological entropy of the plant. We then look into the case of the optimal state estimation over channels with multiplicative noise. By a sequential design under the channel/estimator co-design framework, we show that the estimation error covariance of the optimal estimator is bounded under channel resource allocation if the total channel capacity is greater than the topological entropy.