Moderate Deviations of the Random Riccati Equation
Abstract
We characterize the invariant filtering measures resulting from Kalman filtering with intermittent observations (\cite{Bruno}), where the observation arrival is modeled as a Bernoulli process. In \cite{Riccatiweakconv}, it was shown that there exists a $\overline{\gamma}^{\{\scriptsize{sb}}}>0$ such that for every observation packet arrival probability $\overline{\gamma}$, $\overline{\gamma}>\overline{\gamma}^{\{\scriptsize{sb}}}>0$, the sequence of random conditional error covariance matrices converges in distribution to a unique invariant distribution $\mathbb{\mu}^{\overline{\gamma}}$ (independent of the filter initialization.) In this paper, we prove that, for controllable and observable systems, $\overline{\gamma}^{\{\scriptsize{sb}}}=0$ and that, as $\overline{\gamma}\uparrow 1$, the family $\{\mathbb{\mu}^{\overline{\gamma}}\}_{\overline{\gamma}>0}$ of invariant distributions satisfies a moderate deviations principle (MDP) with a good rate function $I$. The rate function $I$ is explicitly identified. In particular, our results show:
 Publication:

arXiv eprints
 Pub Date:
 October 2009
 arXiv:
 arXiv:0910.4686
 Bibcode:
 2009arXiv0910.4686K
 Keywords:

 Mathematics  Probability;
 Computer Science  Information Theory;
 Mathematics  Dynamical Systems;
 Mathematics  Optimization and Control
 EPrint:
 Revised Version, 35 pages