摘要
基于最大容量和最小差错率准则,研究了在GMD V-BLAST系统中的天线选择问题。选用以信道矩阵非零奇异值的几何均值最大化为目标函数,可以避免容量与差错率性能之间的矛盾。在所有可用天线中进行选择的全搜索算法虽有最佳性能,但复杂度太高。基于贪婪算法,对发射天线采用快速的逐增选择策略,对接收天线采用快速的逐减选择策略,可以显著降低计算的复杂性。计算机仿真结果表明,所采用的快速天线选择算法可以较低的复杂度获得接近全搜索法的容量和分集增益。
The antenna selection technology for Geometric mean decomposition (GMD) Vertical Bell lab layered space-time (V-BLAST) systems is studied based on maximizing capacity criterion and minimizing error probability criterion in this paper. Maximizing the geometric mean of the nonzero singular values of the channel matrix as a target function for antenna selection can avoid the contradiction between the capacity and SER performance, by which we adopt it throughout the discussion of selection algorithm. The exhaustive search algorithm that need search for the required antenna subset in all the antennas available has the optimal performance but too high computation complexity that obstructs its applications. To reduce the complexity, two greedy algorithms, adding one antenna to initial empty set step by step for transmitting antenna selection and excluding one antenna from initial set step by step for receiving antenna selection, are presented. Simulation results show that the two fast algorithms can achieve almost the same capacity and SER performance as that of the exhaustive search algorithm with lower computation complexity.
出处
《南京邮电大学学报(自然科学版)》
EI
2008年第4期7-11,共5页
Journal of Nanjing University of Posts and Telecommunications:Natural Science Edition
关键词
天线选择
多输入多输出
几何均值分解
贝尔实验室垂直分层空时
Antenna selection
Multiple input multiple output
Geometric mean decomposition
Vertical Bell lab layered space-time