多输入多输出单载波频分多址系统的鲁棒波束赋形算法

发布时间：2018-03-04 00:03

本文选题：波束赋形　切入点：多输入多输出　出处：《西安交通大学学报》2017年08期 　论文类型：期刊论文

【摘要】：针对多输入多输出(MIMO)单载波频分多址(SC-FDMA)系统中现有波束赋形算法在非理想信道状态信息下性能下降以及对信道信息误差敏感的问题,提出了一种鲁棒波束赋形算法。该算法利用谱范数对信道状态信息误差进行建模,并基于均方误差和准则将波束赋形优化问题建模为一个极小极大化问题,然后利用优超理论以及奇异值不等式理论证明该优化问题的鞍点特性并得出最优的波束赋形结构,最后利用凸优化得到最优的功率分配算法。仿真结果表明:与非鲁棒波束赋形和等功率分配算法相比,该鲁棒波束赋形算法在误码率以及频谱效率方面都能获得更好的性能,信道状态信息误差增大时,该鲁棒波束赋形算法与非鲁棒波束赋形算法相比在高信噪比时误码率性能约有1dB增益,频谱效率约提升6%。
[Abstract]:In this paper, the existing beamforming algorithms in single-carrier frequency division multiple access (SC-FDMA-FDMA) systems with multiple input / multiple output (MIMO-MIMOA) systems have poor performance and sensitivity to channel information errors under the condition of non-ideal channel state information. A robust beamforming algorithm is proposed, in which the channel state information error is modeled by spectral norm, and the beamforming optimization problem is modeled as a minimization problem based on the mean square error and criterion. Then the saddle point property of the optimization problem is proved by using the optimal supertheory and the singular value inequality theory, and the optimal beamforming structure is obtained. Finally, the optimal power allocation algorithm is obtained by convex optimization. The simulation results show that compared with the non-robust beamforming and equal power allocation algorithms, the robust beamforming algorithm can achieve better performance in both BER and spectral efficiency. When the channel state information error increases, the BER performance of the robust beamforming algorithm is about 1 dB gain and the spectral efficiency is about 6 times higher than that of the non-robust beamforming algorithm at high signal-to-noise ratio (SNR).
【作者单位】：哈尔滨工业大学电子与信息工程学院;
【基金】：国家自然科学基金资助项目(61671179) 国家“973计划”资助项目(2013CB329003)
【分类号】：TN929.5

【参考文献】