A Near Optimal Antenna Assignment for MIMO Systems with Low Complexity

High performance, low complexity schedulers are challenging to develop for the downlink of multiple-input multiple-output (MIMO) cellular systems. Selection among available channels between users and the base station in a centralized way makes the problem complex. An approach to solve this problem is combinatorial optimization algorithms. In this case, the scheduling can be done in two different phases. The first phase is the user selection in which the scheduler selects a group of users based on its own criterion. The assignment scheme assigns the selected users to the transmit antennas, taking into consideration the capacity maximization in the second phase of the scheduling. Considerations related to user satisfaction, fairness, physical layer parameters, and/or traffic arrival processes in this kind of scheduler can be implemented with low complexity. For the second phase, we propose a near-optimal, low-complexity, assignment scheme, called the max difference of top two (MDTT). An intensive simulation study taking into account the mobility of users shows the superior performance of our assignment scheme.