基于YHFT-Matrix的MIMO-OFDM系统关键算法的设计与实现

发布时间：2019-06-12 11:20

【摘要】：随着新一代的LTE、4G无线通信技术的迅猛发展，MIMO-OFDM系统成为国内外的研究热点。无线通信技术高速的发展，导致了多种通信协议和标准并存，这成为了无线通信进一步发展的阻碍。软件无线电技术是解决上述问题的有效方法之一，该方法依托数字信号处理器，以软件方式实现无线通信系统中的各种功能模块。传统的标量单核处理器难以满足软件无线电应用的高密度实时计算需求，基于向量多核处理器实现软件无线电成为研究的主流方向。 YHFT-Matrix是国防科学技术大学自主研发的一款面向软件无线电的高性能向量多核处理器，针对YHFT-Matrix体系结构特点，软件实现高效的MIMO-OFDM系统具有重要的理论意义和应用价值。本文主要的工作包括：（1）深入研究了4G通信系统中的MIMO-OFDM技术，在分析其系统模型的基础上，提出一种基于YHFT-Matrix平台的MIMO-OFDM系统设计与实现方法；（2）设计与实现了基于YHFT-Matrix平台的大、小矩阵乘法运算向量化的方法，该方法充分利了YHFT-Matrix对数据向量处理的能力。实验结果表明，对多个4×4矩阵相乘、一个4×4矩阵同时与多个4×4矩阵相乘和规模大于16的矩阵相乘的运算，取得了高效的计算性能，同比TIC62X和TIC64X系列芯片，分别获得了9.4和4.9以上的加速比；（3）深入研究MIMO-OFDM系统中STBC编码技术，设计和实现了基于YHFT-Matrix平台的2×2和4×4系统的STBC编码运算方案，，实验结果表明，针对同样的算法功能，同比TIC67X，该方法获得了8倍的性能加速；（4）深入研究MIMO-OFDM系统中的均衡算法，分别设计和实现了基于YHFT-Matrix平台的1×1、2×2和4×4三种天线系统的均衡算法，对2×2、4×4系统分别采用了快速求逆算法和基于分块矩阵的求逆的方法作映射实现。实验结果表明，YHFT-Matrix对均衡运算提供了高效的支持，其中4×4系统的MIMO均衡运算性能相比TIC67X取得40倍的加速。
[Abstract]:With the rapid development of the new generation of LTE,4G wireless communication technology, MIMO-OFDM system has become a research focus at home and abroad. The rapid development of wireless communication technology has led to the coexistence of a variety of communication protocols and standards, which has become an obstacle to the further development of wireless communication. Software radio technology is one of the effective methods to solve the above problems. This method relies on digital signal processor to realize various functional modules in wireless communication system by software. The traditional scalar single-core processor is difficult to meet the high-density real-time computing requirements of software radio applications. The implementation of software radio based on vector multi-core processor has become the mainstream research direction. YHFT-Matrix is a high performance vector multi-core processor for software radio developed by National Defense University of Science and Technology. According to the characteristics of YHFT-Matrix architecture, software implementation of efficient MIMO-OFDM system has important theoretical significance and application value. The main work of this paper is as follows: (1) the MIMO-OFDM technology in 4G communication system is deeply studied, and on the basis of analyzing its system model, a design and implementation method of MIMO-OFDM system based on YHFT-Matrix platform is proposed; (2) the vector quantification method of large and small matrix multiplication operation based on YHFT-Matrix platform is designed and implemented, which fully benefits the ability of YHFT-Matrix to process data vector. The experimental results show that for multiple 4 脳 4 matrices, one 4 脳 4 matrix is multiplied by multiple 4 脳 4 matrices at the same time and the matrix larger than 16 is multiplied at the same time, and the efficient computational performance is obtained. The acceleration ratios of TIC62X and TIC64X series chips are more than 9.4 and 4.9, respectively. (3) the STBC coding technology in MIMO-OFDM system is deeply studied, and the STBC coding operation scheme of 2 脳 2 and 4 脳 4 systems based on YHFT-Matrix platform is designed and implemented. The experimental results show that for the same algorithm function, the performance of this method is 8 times faster than that of TIC67X,. (4) the equalization algorithm in MIMO-OFDM system is deeply studied, and the equalization algorithms of 1 脳 1, 2 脳 2 and 4 脳 4 antenna systems based on YHFT-Matrix platform are designed and implemented respectively. the fast inverse algorithm and the inverse method based on block matrix are used to realize the 2 脳 2 and 4 脳 4 systems, respectively. The experimental results show that YHFT-Matrix provides efficient support for equilibrium operation, in which the performance of MIMO equilibrium operation of 4 脳 4 system is 40 times faster than that of TIC67X.
【学位授予单位】：国防科学技术大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP332;TN929.53

【参考文献】