Multiuser MIMO User Selection Based on L2-Hausdorff Distance with Block Diagonalization

Block diagonalization (BD) is an important precoding method for multiuser multiple-input and multiple-output (MU-MIMO) broadcast channel (BC) systems. When the number of users is large, user selection (scheduling) should be performed so as to make full use of BD precoding. In this paper, based on $$L_{2}$$L2-Hausdorff distance, we propose two novel user selection algorithms, namely GUS-$$\hbox {D}_{{L_{2}}}$$DL2 and US-$$\hbox {S}_{{L_{2}}}$$SL2, for MU-MIMO BC systems with BD precoding to maximize the throughput. Both of the proposed algorithms select users iteratively and perform power allocation using the waterfilling method. In GUS-$$\hbox {D}_{{L_{2}}}$$DL2, $$L_{2}$$L2-Hausdorff distance is used as a crucial factor for the user selection criterion. In US-$$\hbox {S}_{{L_{2}}}$$SL2, besides the criterion in GUS-$$\hbox {D}_{{L_{2}}}$$DL2, a similarity criterion based on $$L_{2}$$L2-hausdorff distance is employed to reduce the cardinality of the candidate user set effectively. The complexities of the proposed algorithms are analyzed and compared with those of typical existing algorithms. Simulations have been carried out to verify the performance of the proposed algorithms. Numerical results suggest that the proposed algorithms outperform most of their rivals in terms of complexity and show competitive performance in throughput.