多基站多用户MIMO系统下行链路基于准最大化和速率预编码和功率分配的双层交替迭代结构

Alternating iterative structure between quasi-maximum sum-rate precoder and power allocation for multi-cell multiuser MIMO downlinks

在多基站多用户MIMO系统中,最大化和速率(maximizing sum-rate,MSR)预编码是一种线性预编码,但是无闭式解.本文提出了一种交替迭代结构的准最大化和速率(Quasi-MSR)预编码方案来逼近MSR,将MSR的求解问题转化成最大化二次分式函数的乘积,并且设计了一种单接收天线用户环境下基于最大化和速率准则的低复杂度交替网格搜索功率分配(alternating grid search power allocation,AGSPA)方案.在此基础上,本文又提出了基于Quasi-MSR和低复杂度AGSPA的双层交替迭代结构(alternating iterative structure,AIS).仿真结果表明:考虑大尺度衰落和天线相关性时,AIS结构在和速率性能上明显优于4种典型的预编码算法(最大化信泄噪比、块对角化、最小均方误差与速率最大化).

In multi-cell multiuser MIMO systems, the precoder of maximizing sum-rate （MSR） is optimal in linear and sum rate senses. However, there is no closed-ibrm solution for MSR. In this paper, all alternating itcrative quasi-MSR precoder of approaching MSR is developed by transforming the MSR problem into the one maximizing the product of fractional quadratic functions. In addition, a low-complexity alternating grid search power allocation （AGSPA） method is designed based on MSR principle in the case of single receive antenna per user. Then, an alternating iterative structure （AIS） between the proposed quasi-MSR and the low-complexity AGSPA is proposed. From simulations, the proposed AIS performs much better than the four common precoders maximizing signal-to-leakage-and-noise ra.t.io, block diagonalization, minimum mean square error, and rate maxi- mization in terms of stun rate in the presence of large-scale fading and antenna correlation.