Efficient Margin Adaptive Scheduling for MIMO-OFDMA Systems

In this paper we address the problem of margin adaptive scheduling in the downlink of an orthogonal frequency division multiple access (OFDMA) multiple-input multiple-output (MIMO) system. Optimal resource allocation in MIMO systems requires the joint optimization of: a) linear transmit and receive spatial filters, b) channel assignment and c) power allocation. This problem is not convex and its complexity becomes thus intractable already for small sets of users and subcarriers. To reduce the complexity of the problem at hand, we propose a novel heuristic strategy that partitions the users in different groups according to their average channel quality and addresses the original problem by solving a succession of lower-complexity allocation problems. The spatial dimension is employed to prevent multiple access interference from hindering the performance of the sequential allocation. To further reduce the complexity burden we introduce a linear programming formulation in combination with a waterfilling-based strategy to allocate channels and power to the groups of users. Numerical results and evaluation of the computational complexity show that, though suboptimal, in most cases the proposed algorithm manages to exploit in an original way the inherent multi-user diversity of multi-carrier systems to ease the task of resource allocation with a very limited performance loss from the theoretic optimum.