Paper detail

Exponential stability and asymptotic properties of the optimal filter for signals with deterministic hyperbolic dynamics

The problem of stability of the optimal filter is revisited. The optimal filter (or filtering process) is the conditional probability of the current state of some stochastic process (the signal process), given both present and past values of another process (the observation process). Typically the filtering process satisfies a dynamical equation, and the question investigated here concerns the stability of this dynamics. In contrast to previous work, signal processes given by the iterations of a deterministic mapping $f$ are considered, with only the initial condition being random. While the stability of the filter may emerge from strong randomness of the signal processes, different and more dynamical effects will be exploited in the present work. More specifically, we consider uniformly hyperbolic $f$ with strong instabilities providing the necessary mixing. This however requires that the filtering process is initialised with densities exhibiting already a certain level of smoothness. Furthermore, $f$ may also have stable directions along which the filtering process will eventually not have a density, a major technical difficulty. Further results show that the filtering process is asymptotically concentrated on the attractor and furthermore will have densities with respect to the invariant (SRB)~measure along instable manifolds of $f$.