Multiuser Scheduling in HSDPA using Simulated Annealing

In this paper, the issue of multiuser scheduling in the context of the high speed downlink packet access (HSDPA) is addressed. Given limited radio resources at the base station, the main challenge is to assign these resources to multiple users at each scheduling period. To perform efficient assignment, channel state information (CSI) for users is required, which can be obtained via their respective feedback channels. Multiuser scheduling is studied in (Kwan and Leung, 2007), in which the availability of perfect CSI is assumed. However, such an approach incurs a high bandwidth overhead. In this paper, a more realistic imperfect CSI feedback is assumed, which takes the form of a finite set of channel quality indicator (CQI) as specified in the HSDPA standard (Holma and Toskala, 2006). A global optimal integer programming and a simulated annealing approaches are proposed in solving the optimization problem. Simulation results show that the performance of the latter approach is very close to that of the former at a much lower complexity. On the other hand, the performance due to a simple greedy approach is significantly worse.