Aspect Angle Classification via Physically Realizable Matched Illumination Waveform Design

Matched illumination waveform design seeks to improve classification and detection performance by leveraging a priori information about the expected scattering from the illuminated environment. This approach leads to a waterfilling solution, in which more power is allocated to high-SNR frequency channels, resulting in improved discrimination capabilities. While much research has been devoted to the derivation and design of matched illumination waveforms, the question remains of how much benefit is obtained by the matched illumination approach. In the context of physical waveforms, we compare the classification and detection performance of matched illumination waveforms to that of traditional optimized RFM waveforms. The relative capability of MI is examined via Monte Carlo simulation, where the simulated test object consists of simulated azimuthdependent scattering characteristics of a Toyota Camry.