Parallel implementations of Kalman filters for tracking applications onto transputer networks

A coordinate system frame is described for manoeuvring target motion. Using the Cartesian coordinate frame system, target motion is described in a simple form with the measurements in polar coordinates. An extended Kalman filter (EKF) or second-order Gaussian filter (SOGF) is used according to the frame reference to estimate the kinematic components of the manoeuvring targets (i.e. position, velocity and acceleration) from noisy measurements. The paper derives the sequential EKF and SOGF tracking algorithms when the range rate measurements are available and investigates the parallel implementation for these algorithms onto a range of transputer topologies to enable practical realisations. Real-time implementation results are presented and the relative speedup and efficiency are calculated to evaluate the performance of each parallel implementation