Parametric UWB radar target modeling using Particle Swarm Optimization

This paper uses the particle swarm optimization for parametrically modeling the impulse response of radar targets segment by segment. Parametric modeling of UWB radar signatures has been used for target recognition [Carriere and Moses, 1992 and 1988]. The performance of the modeling techniques used in this paper is comparable to traditional modeling schemes such as the Prony method, but makes no assumptions about signal statistics, and most importantly does not require a priori knowledge of model order (number of scatterers) but adaptively converges to the proper order.