Low-complexity user selection strategies in the downlink of multi-user channels

In this paper, we propose a new low-complexity user selection algorithm for multi-user downlink channels. We only consider the scenario that linear beamforming, such as zero-forcing beamforming, is employed at the base station (BS). In this scenario, the number of concurrent active users is limited by the number of transmit antennas at the BS. Thus, there arises the problem how to select the optimal set of active users that can maximize the downlink channel capacity based on the channel state information (CSI) of all users. However, the optimal solution to this problem is an exhaustive search of all possible combinations of the users, which has prohibitive computational complexity. We propose a sub-optimal greedy user selection algorithm based on the norm and correlation between the channel matrices of different users. Our new algorithm requires low computational complexity and provides nearly optimal performance. Simulation results are presented to illustrate the good performance of our algorithm