全双工大规模MIMO异构网络频谱效率分析及优化

Spectral Efficiency Analysis and Optimization for Full Duplex Heterogeneous Network with Massive MIMO

摘要: 本文研究全双工大规模多输入多输出（multiple-input-multiple-output, MIMO）异构网络。推导了采用迫零（zero-forcing, ZF）预编码/检测方案的上下行渐近信干噪比（signal-to-interference-plus-noise ratio, SINR）的闭合表达式。根据理论分析，提出了基于几何规划（geometric programming, GP）的功率控制方案和基于部分用户选择（partial users selection, PUS）的跨层干扰消除方案，改善系统频谱效率。仿真结果表明，与传统方案相比本文提出的功率控制方案和跨层干扰消除方案能够显著提升异构网络的频谱效率。

Abstract: In this paper, a full duplex (FD) multiple-input multiple-output (MIMO) heterogeneous network (HetNet) was researched. First, the s of asymptotic uplink/downlink signal-to-interference-plus-noise ratio (SINR) were derived in closed form with zero forcing (ZF) precoding/detecting method. Then, a geometric programming (GP) based power allocation scheme and a cross-layer interference (CLI) cancelation scheme which selects partial users to suppress CLI were proposed to enhance the spectral efficiency (SE) of the HetNet. Simulation results show that the proposed two schemes achieve better performance when compared with conventional scheme.