RMMSE雷达脉冲压缩快速算法中矩阵求逆的FPGA实现  

FPGA implementation of matrix inversion in RMMSE radar pulse compression express algorithm

作　　者：宋一鑫[1] 姚振东[1] 

机构地区：[1]成都信息工程学院电子工程学院,四川成都610225

出　　处：《成都信息工程学院学报》2009年第5期435-439,共5页Journal of Chengdu University of Information Technology

摘　　要：在FPGA平台上,实现了RMMSE快速算法中三维矩阵的求逆。求逆功能模块基于并行、高效的心动阵列结构,采用上下三角(LU)分解求逆法。它由LU分解模块、三角矩阵求逆模块和矩阵乘积模块串联组成。矩阵数据顺序流过阵列,便得到求逆结果,仅花费19个时钟周期,具有很高的计算效率。将QuartusⅡ仿真结果,与MATLAB计算结果进行了比较,验证了算法的正确性。The three-dimensional matrix inversion is implemented in the RMMSE express algorithm based on a FPGA platform. The inverse function module based on the parallel and efficient systolic array structure uses the LU factor- ization of the matrix inversion method. This module is composed of three part series： LU factorization module, trian- gular matrix inversion module and matrix multiplication module. The matrix data flow through the array in accor- dance with a high computational efficiency. The comparison between the QuartuslI simulation results and Matlab cal- culation results shows the validity of this algorithm.

关 键 词：RMMSE 矩阵求逆 LU分解 FPGA 

分 类 号：TN958.3[电子电信—信号与信息处理]