Efficient filter implementation using QRD-RLS algorithm for Phased Array Radar applications

This paper illustrates the two different approaches for implementation of an adaptive filter using QRD-RLS (Q R decomposition based Recursive Least Squares) algorithm for Phased Array Radar application. One approach involves back substitution procedure whereas another involves updating of inverse data matrix. The former approach is called as Conventional QRD-RLS and the later is called as Inverse QRDRLS. QRD-RLS algorithm is numerically more stable when compared to traditional algorithms such as LMS and conventional RLS. Also QRD-RLS algorithm is suitable for parallel and pipelined implementation thus making it very useful in the applications where speed, accuracy and numerical stability are of utmost importance; such as Phased Array Radar receiver. A high performance FPGA Virtex 6vlx240t is employed to implement the above mentioned algorithms.