大规模MIMO系统分级预编码及基于人工免疫的功率分配算法

Hierarchical Precoding and Artificial Immune-based Power Allocation Algorithm for Massive MIMO System

现有大规模多输入多输出(Multiple Input Multiple Output,MIMO)系统通常采用单一的预编码方案,进行功率分配时往往不能兼顾系统整体和用户个体的性能,且传统求解方法计算复杂度高。针对上述问题,首先提出一种分级预编码策略,依据用户的位置信息在小区内将用户进行分组以消除多用户干扰(Multi-user Interference,MUI),并基于信漏噪比设计预编码矩阵抑制小区间干扰(Inter-cell Interference,ICI)以及残存MUI;其次将功率分配的优化目标函数设计为每小区用户最小频谱效率与平均频谱效率的乘积;最后提出了基于人工免疫算法的功率分配方法。实验结果表明,与传统的功率分配算法相比,所提算法在50%的中断概率点处,在保证Maxmin fairness的同时,使得用户的平均频谱效率提高了约29%。

The existing Massive MIMO(Multiple Input Multiple Output)systems usually use a single precoding scheme,which often cannot consider the performance of the whole system and individual users when performing power allocation,and the traditional solutions have a high computational complexity.To address the above problems,this paper first proposes a hierarchical precoding strategy,which groups users in cells according to their location information to eliminate MUI(Multi-User Interference),and designs a precoding matrix based on the signal leakage noise ratio to suppress ICI(Inter-Cell Interference)and residual MUI.Then,it designs the optimization objective function of power allocation as the product of the minimum spectral efficiency and the average spectral efficiency of each cell user.Finally,this paper puts forward a power allocation method based on artificial immune algorithm.The experimental results indicate that,compared with traditional power algorithms,the proposed algorithm increases the average spectral efficiency of users by about 29%while ensuring max-min fairness at the point of 50%outage probability.