High-frequency statistical classification of complex targets using severely aspect-limited data

Many existing methods for asymptotic inverse scattering rely heavily on simplifying assumptions such as convex target structure and full-aspect data sets. As applied to real-world scattering situations (for example, radar target identification) these assumptions are rarely justified. We investigate a statistical method appropriate to very complex targets consisting of many scatterering centers, each scatterer with different morphological properties. Under the assumption that the data are collected from a very narrow range of aspects and using only several closely spaced frequencies, a simple calculation shows that gross target size and shape properties are determinable.