User satisfaction based joint user selection and beamforming in TD-LTE-A downlink.

In TD-LTE-A downlink, user selection is a key technique when the number of users exceeds the amount could be supported by eNodeB at one Transmission Time Interval (TTI). In order to improve the fairness of different users and consider the difference of user traffic in the process of user selection, a Signal to Leakage and Noise-Ratio (SLNR) beamforming based joint user selection and beamforming algorithm is proposed in this paper. Since in TD-LTE-A systems, the traffic can be classified into two types: Guaranteed Bit Rate (GBR) traffic and Non-Guaranteed Bit Rate (Non-GBR) traffic, user satisfaction for GBR and Non-GBR traffic are defined at first, and then, based on the difference of traffic, a user satisfaction based joint user selection and beamforming algorithm is proposed. Simulation results show that compared with conventionally used joint user selection and beamforming algorithm, user satisfaction of GBR traffic can be increased by 300% due to the fact that GBR traffic always has higher priority than Non-GBR traffic. Moreover, user fairness can also be improved in terms of the variance of user satisfaction.