| 摘要: |
| 在大规模多输入多输出（Multiple-Input Multiple-Output，MIMO）系统上行链路检测算法中，最小均方误差（Minimum Mean Square Error，MMSE）算法可取得近似最优的性能，然而MMSE算法涉及高维矩阵求逆问题，其计算复杂度高达O(K3)，其中K表示用户数。为此，针对极化信道编码的大规模MIMO系统，基于无转置极小残差（Transpose-Free Quasi-Minimal Residual，TFQMR）方法，提出了一种低复杂度次优信号检测算法。该算法有效地避免了矩阵求逆运算，使其计算复杂度降至约O(K2)。仿真结果表明，基于TFQMR的信号检测算法的误比特率性能与计算复杂度均优于基于Neumann级数展开的信号检测算法；同时，最多经5次迭代该方法可取得接近MMSE检测算法的性能。 |
| 关键词: 大规模MIMO 信号检测 最小均方误差 无转置极小残差 |
| DOI： |
|
| 基金项目:国家自然科学基金资助项目（61701062）；重庆市基础与前沿研究计划项目（cstc2019jcyj-msxmX0079） |
|
| A low complexity signal detection algorithm based on TFQMR for massive MIMO systems |
| CHEN Hongyan,LI Gang,JING Xiaorong |
| (1.School of Communication and Information Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;2.ZTE Corporation,Shenzhen 518000，China) |
| Abstract: |
| Minimum mean square error(MMSE) detection algorithm can achieve near optimal performance for uplink massive multiple-input multiple-output(MIMO) systems.However,the MMSE algorithm needs the calculation of the inversion of the high-dimension matrix,which involves the computational complexity as high as O(K3) with K denoting the number of users.Therefore,for the polar-channel-coded massive MIMO systems,a low-complexity sub-optimal signal detection algorithm is proposed based on the transpose-free quasi-minimal residual(TFQMR) method.The proposed algorithm avoids matrix inversion and reduces its computational complexity to O(K2).Simulation results shows that the proposed TFQMR-based signal detection algorithm outperforms the Neumann series expansion algorithm on both the bit error rate performance and computational complexity.Meanwhile,the TFQMR-based algorithm converges and approaches the performance of the MMSE detection algorithm after at most five iterations. |
| Key words: massive MIMO signal detection minimum mean-square error transpose-free quasi-minimal residual |
