FPGA实现基于施密特正交化的自适应算法

QRD-SMI Algorithm Based on MGS Algorithm and its FPGA Implementation

在自适应波束形成算法理论基础上,将修正施密特正交化算法运用在基于三角分解的样本矩阵求逆算法中,提出了基于修正施密特正交化算法的三角分解矩阵求逆算法,并给出了具体的可编程门阵列硬件实现结构。该硬件结构的实现采用了坐标旋转数字计算机代替该算法中的除法运算,有效地节约了硬件资源消耗和结构时延。通过硬件仿真对计算结果进行分析研究,实验证明该硬件结构不仅有良好的数值稳定性,能对干扰有效抑制,且硬件资源消耗少,高度模块化。

Based on the theoretical analysis of adaptive beamforming algorithm, the modified Gram Schmidt or- thogonal algorithm （MGS） is used in the algorithm based on the QR decomposition of inversion sample matrix （ QRD- SMI） to propose the QRD-SMI algorithm based on the MGS algorithm with specific programmable gate array （FPGA） implementation structure given： multiplexing QR decomposition array （MQRD-IMGS） based MGS. The FPGA im- plementation structure uses coordinate rotation digital computer （CORDIC） to replace the division operational in MGS, effectively saving hardware resource consumption and the structure time delay. Hardware simulation shows that the highly modular MQRD-IMGS maintains numerical stability with good suppression of the interfering signals and small hardware resource consumption.