User Scheduling for Heterogeneous Multiuser MIMO Systems: A Subspace Viewpoint
In downlink multiuser multiple-input multiple-output (MU-MIMO) systems, users are practically heterogeneous in nature. However, most of the existing user scheduling algorithms are designed with an implicit assumption that the users are homogeneous. In this paper, we revisit the problem by exploring the characteristics of heterogeneous users from a subspace point of view. With an objective of minimizing interference non-orthogonality among users, three new angular-based user scheduling criteria that can be applied in various user scheduling algorithms are proposed. While the first criterion is heuristically determined by identifying the incapability of largest principal angle to characterize the subspace correlation and hence the interference non-orthogonality between users, the second and third ones are derived by using, respectively, the sum rate capacity bounds with block diagonalization and the change in capacity by adding a new user into an existing user subset. Aiming at capturing fairness among heterogeneous users while maintaining multiuser diversity gain, two new hybrid user scheduling algorithms are also proposed whose computational complexities are only linearly proportional to the number of users. We show by simulations that the effectiveness of our proposed user scheduling criteria and algorithms with respect to those commonly used in homogeneous environment.