Impact of Radar Target Covariance Structure on SINR Maximization

The target response in radar systems is often modeled as either deterministic and known, or as a random process. This can be justified in that uncertainties in target aspect angle and hence RCS pattern can make more detailed assumptions of the target response difficult. This leads to the common simpler models of RCS based on statistical descriptions as random variables with some probability density function. However, the ability to simulate radar cross section (RCS) data over a wide range of frequencies and aspect angles allows for a high dimensional representation of targets that can provide opportunity to go beyond these simplifying assumptions. This paper presents one such approach, where this higher dimensional knowledge can lead aspect angle invariant representations of the targets. Some implications of this are derived in the form of signal-to-interference-plus-noise ratio maximizing receive filters and waveforms.