The Air Force's need to identify hostile threats from beyond visual range is an essential goal if air superiority is to be maintained into the next century. One promising avenue of non-cooperative identification lies in the characterization of physical attributes associated with a particular threat. Such physical attributes can be derived from reception and analysis of radar pulses emitted from hostile platforms. Accordingly, we address the problem of the passive identification of radar emitters in emitter dense and electromagnetically cluttered environments. We propose to develop novel non-orthogonal wavelet representations to characterize specific emitters as an alternative to currently applied Fourier methods. In addition, we present clustering schemes to perform the identification and formulate performance measures by which to judge competing methods.
Keywords: Emitter Id Nonorthogonal Identification Signal Processing Wavelet Radar Representation
