Mean-square asymptotic analysis of cross-coupled Kalman filter state-estimation algorithm for bilinear systems

In this paper, we present an asymptotic analysis of a recursive cross-coupled Kalman filter algorithm for estimating the state of a partially observed bilinear stochastic system. The cross-coupled Kalman filter algorithm consists of two Kalman filters-each Kalman filter estimating the state of one of the two state components of the bilinear system. Our asymptotic analysis involves mean square asymptotic results on the tracking capabilities of the resulting cross-coupled Kalman filter algorithm.