Detection architecture with improved classification capabilities for covariance structures

This paper proposes a radar adaptive detection architecture composed of an interference covariance matrix structure classifier before a bank of adaptive radar detectors. The former relies on the model order selection framework. This classifer accounts for six interference covariance matrix structure classes, including the additional Toeplitz structures. The proposed architecture exhibits an improved estimation of covariance matrix, and better detection performance with an improved classifer, especially in the presence of a small volume of training samples. As a consequence, the newly proposed architecture can guarantee excellent detection performance for a wider class of operating scenarios. Numerical examples show that the proposed architecture exhibits improved detection performance with respect to its competitor when the interference covariance matrix exhibits the Toeplitz structure. (c) 2022 Elsevier Inc. All rights reserved.

Automated summary

This paper proposes a radar adaptive detection architecture with an interference covariance matrix structure classifier and a bank of adaptive radar detectors. The architecture improves the estimation of the covariance matrix and the detection performance, especially with a small volume of training samples. Therefore, the newly proposed architecture can guarantee excellent detection performance for a wider class of operating scenarios.