Multi-perspective target classification

Traditional radar classification is based on a single aspect view of the target. It is remarkably sensitive to the particular direction under which the target is sensed. A network of radars offers means of collecting backscattered radiation of a range of viewing angles. This offers multi-perspective data sets as an aid to improve radar target classification. By definition, this will provide a data set with higher information content. In this paper we examine whether this approach is useful for classification purposes. Results using full scale turntable demonstrate significant improvements in performance over and above the monostatic case.