Singular Value Decomposition Algorithm of Rectangular Complex Matrix Based on FPGA
-
摘要: 在OFDM和MIMO系统中普遍使用长方形矩阵复数奇异值分解运算。针对传统算法运算量大,迭代次数多的问题,提出了一种基于householder和双边Jacobi的混合优化算法。该算法首先通过householder变换将矩阵化解为二对角矩阵;然后提取2×2复矩阵;再进行改进型复数双边Jacobi变换。兼具有QR算法的高精度和Jacobi算法的低硬件实现成本的优点。给出了2×8的CSVD的FPGA硬件实现方案并进行了板级测试。测试结果表明,该混合优化算法较传统算法在硬件资源上节省26%,延时缩短10倍,在同等位宽下计算精度至少提高了一个数量级。
-
关键词:
- 复数奇异值分解 /
- 可编程逻辑阵列 /
- householder /
- Jacobi /
- 长方矩阵
-
[1] AU E K S, JIN S, MCKAY M R, et al. Analytical performance of MIMO-SVD systems in ricean fading channels with channel estimation error and feedback delay[J]. IEEE Transactions on Wireless Communications, 2008, 7(4): 1315-1325. [2] ZHAO Xue-zhi, YE Bang-yan. Convergence characteristic of single direction shrink QR algorithm in the singular value decomposition[J]. Journal of University of Electronic Science and Technology of China, 2010, 39(5): 762-767. [3] SRINIVASAN J, RAJARAM S. FPGA implementation of precoding using low complexity SVD for MIMO-OFDM systems[C]//Information Communication and Embedded Systems (ICICES). [S.l.]: IEEE, 2013. [4] CHAKROBORTY S, SAHA G. Feature selection using singular value decomposition and QR factorization with column pivoting for text-independent speaker identification [J]. Speech Communication, 2010, 52(9): 693-709. [5] 胡谋法, 董文娟, 王书宏, 等. 奇异值分解带通滤波背景抑制和去噪[J]. 电子学报, 2008, 36(1): 111-116. HU Mou-fa, DONG Wen-juan, WANG Shu-hong, et al. Singular value decomposition band-pass-filter for image background suppression and denoising[J]. Acta Electronica Sinica, 2008, 36(1): 111-116. [6] WANG Y, CUNNINGHAM K, NAGVAJARA P, et al. Singular value decomposition hardware for MIMO: State of the art and custom design[C]//Reconfigurable Computing and FPGAs (ReConFig). [S.l.]: IEEE, 2010. [7] HAN Q, ZENG L. FPGA Implementation for low-rank channel estimation of OFDM[J]. Journal of Networks, 2012, 7(10): 1631-1638. [8] HEMKUMAR N D, CAVALLARO J R. A systolic VLSI architecture for complex SVD[C]//Circuits and Systems, ISCAS'92. [S.l.]: IEEE, 1992. [9] 赵学智, 叶邦彦. 单向收缩QR算法在奇异值分解中的收敛特性[J]. 电子科技大学学报, 2010, 39(5): 762-767. [10] MA W, KAYE M E, LUKE D M, et al. An FPGA-based singular value decomposition processor[C]//Electrical and Computer Engineering. [S.l.]: IEEE, 2006. [11] LIU J, ZHANG J. A new maximum simplex volume method based on householder transformation for endmember extraction[J]. IEEE Transactions on Geoscience and Remote Sensing, 2012, 50(1): 104-118. [12] PEDRAM A, GERSTLAUER A, GEIJN R A V D. Floating point architecture extensions for optimized matrix factorization[C]//Proceedings of the 2013 IEEE 21st Symposium on Computer Arithmetic. [S.l.]: IEEE, 2013: 49-58. [13] 张贤达. 矩阵分析与应用[M]. 北京: 清华大学出版社有限公司, 2004. ZHANG Xian-da. Matrix analysis and applications[M]. Beijing: Tsinghua and Springer Publishing House, 2004.
点击查看大图
计量
- 文章访问数: 5217
- HTML全文浏览量: 169
- PDF下载量: 104
- 被引次数: 0