On the optimal processor assignment for computing the steady state kalman filter in parallel and distributed systems

Abstract

Kalman filters have many practical applications in various fields such as sensor networks, image and video processing. Therefore, their fast computation is of paramount importance. In this paper distributed implementations for the steady state Kalman filter are proposed. The distributed algorithms are based on partitioning the measurement vector, the state vector or both of them. The number of processors is determined a priori. The optimal distribution of measurements/ states into parallel processors minimizing the computation time is also a priori determined. © Springer International Publishing AG 2017.