一种自适应阵列信号处理算法的高速实现

The high-speed realization of an adaptive array algorithm

阐述了一种自适应抗干扰阵列信号处理系统的高速实现.算法采用基于QR分解的最小二乘算法,在系统的设计中采用了ARLA技术.ARLA技术通过引入并行机制把串行的自适应阵列信号处理算法转换成了并行的自适应阵列信号处理算法.ARLA技术的采用使基于QR分解的最小二乘算法具有优良的流水粒度.文章阐述了系统的硬件构架,该构架的宏单元可以由能够完成Givens旋转的CORDIC运算单元来充当,而且可以在不损失收敛速度的前提下达到很高的采样速率.

With the development of electronic technology,people must adopt the complicated adaptive algorithm for systems.This paper presents the realization of adaptive least squares array signal processing algorithm based on QR decomposition adopting annihilation_reorder look_ahead(ARLA)technology.The ARLA technology transforms a sequential adaptive array signal processing algorithm into an equivalent concurrent adaptive array signal processing algorithm by creating additional concurrency in the algorithm.The ARLA is employed to develop fine_grain pipelined QR decomposition_based RLS adaptive algorithm.The proposed architectures can be operated at an arbitrarily high sample rate without degrading the convergence behavior and consist of only Givens rotations,which can be scheduled onto CORDIC arithmetic_based processors.Results indicate the QRD-RLS algorithm adopting ARLA technology can come true at a high speed by hardware framework as explained in this paper.