Simulation Study of the Unified Bayesian-Regularization Technique for Enhanced Radar Imaging

In this paper, we intend to present the results of extended simulation study of the family of the radar image (RI) formation algorithms that employ the recently developed and investigated fused Bayesian-regularization (FBR) paradigm for high-resolution reconstruction of the spatial spectrum pattern (SSP) of the wavefield sources distributed in the remotely sensed environment. The FBR methodology is based on the aggregation of the Bayesian minimum risk statistical optimal estimation strategy with the descriptive weighted constrained least squares optimization technique that involves the non trivial a priori information on the desired properties of the SSP to be reconstructed from the actually measured data signals. The advantages of the well designed RI experiments (that employ the FBR-based methods) over the cases of poorer designed experiments (that employ the matched spatial filtering as well as the constrained least squares estimators) are investigated trough the simulation study.