INHOMOGENEOUS PERFORMANCE EVALUATION OF A NEW METHODOLOGY FOR FLUCTUATING TARGET ADAPTIVE DETECTION

The ideality of operating environment of radar systems is extremely scarce while the demand for these systems is growing at a rapid pace. Technology of adaptation is therefore of primary concern in the design of their future strategies. The difficulty in finding a solution based on a single adaptive algorithm to deal with diverse noise environments has led to the development of composite adaptive procedure. Therefore, fusion of particular decisions of the single adaptive variants through appropriate rules provides a better final detection. This paper is intended to analyze the fusion strategy of cell-averaging (CA), order statistics (OS), and trimmed-mean (TM) schemes in heterogeneous environments. The tested target and the spurious ones are assumed to follow [chi square]-distribution with two- and four-degrees of freedom in their fluctuations. A closed form processor performance is derived. The results show that for the heterogeneous operation, this approach is more realistic. Particularly in multi-target situations, it exhibits higher robustness than CA, OS, or TM architecture. Additionally, our results reveal that it exhibits a homogeneous performance outperforming that of the Neyman-Pearson (N-P) detector which is the yardstick in the world of adaptive detection.