Waveform processing-domain diversity and ATR

Classification of targets by radar has proved to be notoriously difficult with the best systems still yet to attain sufficiently high levels of performance and reliability. In this paper we take cues from nature to propose and examine a novel approach to target classification, based on diversity, as applied in the waveform processing domain. In the new approach, data is processed in multiple, different, forms, in parallel. The two forms that we have exploited in this work are the time and space domains. Most classification and Radar image analysis algorithms handle Radar data in the space domain only. Using simulation studies, we first show that phase or k-space data contains additional information. It is also shown that, counter-intuitively, having a sharp spatial Radar image (with reduced side-lobes) in fact worsens classification performance. Lastly, the proposed architecture is validated against a traditional, unitary based classification scheme.