低复杂度QR-RLS自适应均衡算法的实现

Realization of QR-RLS Adaptive Equalization Algorithm with Low Complexity

传统CORDIC算法需要通过查找表和许多乘法器才能实现多种复杂函数的计算,这会导致硬件电路实现复杂,运算速度降低。在向量模式下提出一种改进型CORDIC算法,它不需要查找表和模校正因子,只需通过简单的移位和加减运算就能实现多种复杂函数的计算,从而能够减少硬件的开销,提高运算的性能。最后将该算法应用于QR-RLS自适应均衡算法的Givens旋转中,完成了输入数据矩阵的QR分解,实现了自适应均衡的效果。

Traditional coordinate rotation digital computer（CORDIC） algorithm can realize the calculation of many complex functions by look-up table and many multipliers,which will result in the complexity of hardware circuit and decrease the operation speed.A modified CORDIC algorithm is proposed in rotation mode.It needn＇t look-up table and scale factor correction and can realize the calculation of many complex functions only by simple shift and addition-subtraction operation,which can reduce the cost of hardware and improve operational performance.Finally,the proposed algorithm is applied to the Givens rotation of QR-RLS adaptive equalization algorithm,which completes QR-decomposition of input data matrix and realizes adaptive equalization effect.