Novel User Scheduling Schemes Based on Nonlinear Precoding for Multiuser MIMO Systems

This paper proposes a novel greedy user ordering algorithm for the multiuser multiple-input multiple-output (MIMO) systems employing block diagonal geometric mean decomposition method and Tomlinson-Harashima precoding (THP). Theoretical analysis and computer simulations illustrate its low computation complexity relative to the optimal user ordering achieving by brute search over all the possible ordering permutations resulting in extremely high computation complexity. Meanwhile the bit error rate (BER) performance of the proposed algorithm is very close to the optimal user ordering. Moreover, in order to mitigate the impact of users with smaller sub-channel gains to the whole system's BER performance, a joint preprocessing scheme design of adaptive data streams reduction and greedy user ordering (ADSR-GUO) is proposed. By means of choosing different values for the controlling factor, we can obtain different system sum-rate and BER performance to satisfy different quality-of-service (QoS) requirements.