论文部分内容阅读
分布式大规模多输入多输出(MIMO)系统中,利用时分双工模式的信道互易特性,发送端的接入节点可以根据估计得到的上行信道状态信息来进行下行发送预编码的设计.然而,完整的通信信道还包含收发两端的射频电路.而射频增益的失配,破坏了通信信道的互易性,降低了系统的性能.本文研究分布式大规模MIMO系统中的互易性校准问题,接入节点和用户均配置多根天线,系统采用块对角化预编码.分析了在射频增益失配时系统的可达速率,推导了总体最小二乘算法的优化目标函数表达式.同时为了避免总体最小二乘算法需要特征值分解所带来的较高复杂度,提出了迭代坐标下降校准算法.理论分析和仿真结果表明,本文所提迭代坐标下降校准算法基本达到了总体最小二乘算法的性能,收敛速度快,大大降低了实现的复杂度.
In distributed large-scale multi-input multiple-output (MIMO) system, using the channel reciprocity characteristic of TDD mode, the access node at the sending end can design the downlink sending precoding according to the estimated uplink channel state information.However, The complete communication channel also contains the RF circuits at both ends of the transceiver, while the RF gain mismatch, which undermines the reciprocity of the communication channel and reduces the performance of the system.This thesis studies the problem of reciprocity calibration in distributed large-scale MIMO systems, The access nodes and users are equipped with multiple antennas and the system adopts block diagonalization precoding.The accessibility rate of the system in the case of RF gain mismatch is analyzed and the optimal objective function expression of the overall least squares algorithm is deduced To avoid the complexity of eigenvalue decomposition that the whole least squares algorithm needs, this paper proposes an iterative coordinate descent calibration algorithm.Theoretical analysis and simulation results show that the descent calibration algorithm proposed in this paper has basically reached the total least square algorithm The performance, fast convergence, greatly reducing the complexity of the implementation.