Performance analysis of adaptive algorithms for space‐time adaptive processor (STAP) in phased array radar

Abstract

The most challenging task in phased array radar system is the mitigation or suppression of noise and interferences to enhance the useful signal in radar received signals. Multiple solutions have been proposed by researchers for suppression of unwanted signals or interferences which have highly degraded the overall system performance. Nowadays, space-time adaptive processor (STAP) which is a two-dimensional filtering technique is commonly used in phased array antenna with multiple spatial channels. A detailed study is performed on the well-known adaptive algorithms least mean square (LMS), normalized least mean square (NLMS) and recursive least square (RLS), to suppress the high state sea clutter by updating the adaptive STAP weight vector. However, RLS is considered as the suitable adaptive algorithm in STAP because of its good convergence rate, but has high computational complexity. To overcome this issue, two variants of RLS (QRD-RLS and Inverse QRD-RLS) are also discussed which reduce the computational complexity of the algorithm. MATLAB simulations are done to verify the performance of LMS and RLS algorithms in terms of accuracy and convergence rate. Finally, hardware implementations of STAP-QRD-RLS and STAP-IQRD-RLS adaptive beamformers are done and evaluated in terms of latency, throughput and efficiency. The selected platform is the Virtex-5 field-programmable gate array.