Lightweight User Grouping with Flexible Degrees of Freedom in Virtual MIMO

Virtual MIMO (Multiple Input Multiple Output) groups multiple single-antenna mobile devices to form an antenna array, offering higher degrees of freedom and improved spatial diversity gain as a real MIMO does, yet with much lower costs. In this paper, we focus on the user grouping problem in uplink transmission from multiple single-antenna users to one multiple-antenna base station. State-of-the-art solutions mostly target two single-antenna users, solving a pairing problem. Having more than two uplink users in a grouping has yet to be addressed. Intuitively, a higher number of users in a VMIMO group offers better spectrum efficiency, and thus more throughput gains could be expected; the group dynamics however becomes higher too, making fairness harder to be achieved with reasonable computation overhead. To address these challenges, we present a novel solution that decomposes the VMIMO user grouping into two steps. We lighten the computations in user grouping by using instantaneous signal to noise ratio (SNR) as selection criteria, and combining it with proportional fairness for larger groups of users. Lightweight computation in using instantaneous SNR in our solution allows faster grouping and feasible scheduling for a large number of users, as well as fast decision on the efficiency of the number of users in each group. We have evaluated our solution under different network configurations, and the results demonstrate that it achieves much higher data throughput as compared to existing solutions and also well preserves fairness.