Efficient Combination of MU-MIMO THP and User Scheduling to Achieve Both High System Capacity and Fairness

Abstract

This letter proposes an efficient method for combining MU-MIMO Tomlinson Harashima precoding (THP) with user scheduling techniques to realize fairness among users as well as high system capacity. The concept of the proposed approach is to first adopt semiorthogonal user selection (SUS) for throughput maximization and then apply proportional fairness (PF) to the remaining users, which regains fairness lowered by using SUS for throughput maximization. Moreover, considering that THP adopted as a precoder for PF requires the calculation of ordering and precoding for each user combination, the proposed method adopts linear precoding (LP) instead of THP to relax the computational effort during operation of PF. This approach is inspired by the fact that LP achieves almost the same system capacity as THP with ordering when conducting PF. The effectiveness of the proposed method is demonstrated by comparing it to MU-MIMO THP with SUS or PF in terms of system capacity and fairness among users, by means of computer simulations.