Energy-efficient scheduling under hard delay constraints for multi-user MIMO system

Abstract

This paper addresses the joint optimization of power allocation and user selection for a multiuser multiple-input multiple-output (MU-MIMO) system under hard delay constraints, where each user's information/data is considered to be useful only if it is delivered within a given deadline and no outage is allowed. Our goal is to minimize total energy consumption for delivering each user's data within the given deadline. We propose two optimal energy-efficient scheduling schemes using dynamic programming (DP) with different access methods, one is based on time-division multiple access (TDMA) scheme, the other one is based zero-forcing beamforming (ZFBF) scheme. Simulation results reveal that comparing to the straightforward round-robin (RR) method, the proposed scheduling schemes both can achieve higher energy efficiency, and the optimal scheduling with ZFBF scheme can outperform the scheduling with TDMA scheme. When the required amount of data for each user is small, the performance of TDMA scheme performs nearly as well as the ZFBF scheme, when the amount of data becomes large, the ZFBF scheme can achieve significant higher energy-efficient gain than the TDMA scheme.