Bayesian tracking in UWB radar sensor networks

A UWB multistatic radar system, typically composed of one transmitter and several receivers, is a promising solution for tracking an intruder moving within a given surveillance area, because of its extraordinary localization accuracy and very low probability of intercept. In this paper, a tracking algorithm based on particle filtering for UWB multistatic impulse radio radars is proposed, considering the specific multistatic radar setting, UWB signals properties, and complexity constraints. The performance of the proposed approach is evaluated through numerical results accounting for the spatial configuration of the receivers, propagation effects, the presence of residual clutter, power constraints imposed to UWB transmitted signals, and noise. Results show that the proposed tracking algorithm based on particle filters provides high estimation accuracy compared to conventional Kalman filtering, even at a low signal-to-noise ratio, with a reasonable complexity.