Evaluation of Grouping Strategies for an Hierarchical SDMA/TDMA Scheduling Process

Adding SDMA capabilities to modern wireless communication systems like IEEE 802.16 WiMAX promises high system capacity gains but raises the problem of combining orthogonal medium access schemes like TDMA with the non- orthogonal SDMA. In order to reduce the complexity and to improve the flexibility of a joint SDMA/TDMA scheduler, the scheduling process can be subdivided into two parts. First, a spatial grouping algorithm forms groups of users that can be co-scheduled because of their spatial separability. Then, a group scheduling algorithm allocates resources of the time domain to the spatial groups. This paper focuses on the first part of the hierarchical SDMA/TDMA scheduling process, i.e., the spatial grouping. It outlines different grouping strategies and it analyzes their computational complexity. Their respective grouping performance is evaluated according to a metric introduced in the paper. Finally, a tree-based grouping strategy is recommended based on its low complexity and good performance.