多天线通信系统中一种自适应分组检测算法  

An adaptive group detection algorithm for multi-antennas communication system

作　　者：景小荣[1] 周正中[1] 

机构地区：[1]电子科技大学通信学院,四川成都610054

出　　处：《电波科学学报》2008年第4期769-774,共6页Chinese Journal of Radio Science

基　　金：国防科技预研基金(No.51417030101DZ02)资助

摘　　要：针对多天线通信系统,提出了一种基于信道质量的自适应分组检测算法。该算法首先对信道矩阵作修正的选主列QR分解,然后根据各检测层的后检测信噪比与预先计算并存储在系统中的分组门限相比较的结果,将待检符号矢量分为敏感符号分组和可靠符号分组。对敏感符号分组优先采用具有接近最大似然性能的球形译码算法检测,在消除了敏感符号分组影响后,对可靠符号分组采用低复杂度的逐层检测算法。由于分组门限的设置依据后检测信噪比的统计特性,而后检测信噪比与信道环境密切相关,因此,该算法能够根据信道质量自适应调整敏感符号分组中的符号数目,从而可通过选择分组门限,实现对系统检测性能和复杂度的折中考虑。计算机仿真验证了该算法的有效性。An adaptive group detection algorithm is presented on the basis of the channel quality for multi-antennas communication system. The algorithm first applies the modified QR decomposition with column pivoting to channel matrix, then divides the symbol vector for detection into sensitive symbol subgroup and the reliable symbol subgroup by comparing the post-detection SNRs with the group thresholds, which are computed and stored in system beforehand. The sensitive symbol subgroup is preferentially detected by near-ML sphere decoding algorithm, while the reliable symbol subgroup is detected layer by layer after eliminating the effect of the sensitive symbol subgroup. Since the thresholds are set according to the statistic characteristic of the post-detection SNRs, which are affinitive with the channel environment, the algorithm can adaptively adjust the symbol number in the sensitive symbol subgroup with the channel environment. Consequently, the tradeoff can be achieved between performance and complexity via selecting group thresholds. The computer simulation verifies the validity of the proposed algorithm.

关 键 词：多天线通信系统 后检测信噪比 Householder变换 QR分解 球形译码算法 

分 类 号：TN92[电子电信—通信与信息系统]