Noncooperative Precoding for Massive MIMO HetNets: SILNR Maximization Precoding

TL;DR

This paper introduces a novel SILNR maximization precoding technique for massive MIMO HetNets that jointly optimizes user scheduling, beamforming, and power allocation to mitigate interference and improve spectral efficiency.

Contribution

It proposes a new noncooperative precoding method based on SILNR maximization, with a low-complexity algorithm for local optimization, outperforming existing techniques.

Findings

Significant increase in ergodic spectral efficiencies.

Improved rate distributions per user.

Effective mitigation of inter-cell and intra-cell interference.

Abstract

Massive multi-input multiple-out (MIMO) is a key ingredient in improving the spectral efficiencies for next-generation cellular systems. Thanks to the channel reciprocity, in time-division-duplexing mode, each base station (BS) can acquire local channel state information at the transmitter (CSIT) for a set of users possibly located in adjacent cells. When the small cell BSs equipped with not-so-many antennas are densely deployed with marcrocells, a simple noncooperative MIMO precoding technique using local CSIT fails to achieve high spectral efficiency because of strong inter-cell-interference (ICI). In this paper, we present a novel noncooperative massive MIMO precoding technique called signal-to-interference-plus-leakage-plus-noise-ratio (SILNR) maximization precoding. The key idea of the proposed precoding is to jointly find a scheduled user set per cell, the beamforming vectors for the users, and the allocated power by simultaneously mitigating both inter-user-interference (IUI) and ICI leakage power using local CSIT. To accomplish this, we present a low-complexity algorithm that finds a local-optimal solution of the maximization problem for a lower bound of the sum spectral efficiency, i.e., a non-convex optimization problem. By system-level-simulations, we show that the proposed precoding method considerably outperforms the existing noncooperative precoding techniques in terms of the ergodic spectral efficiencies and rate distributions per user.