In the conventional working way of the cellular networks, the scheduling algorithm, which shares the radio resources among active users, only schedules a single user during a TTI. Actually, it has been shown that this working way is not optimal. When channel quality is poor, it would be better to share resources among multiple users. But this should be done carefully, to limit interference between each other. There are many scheduling algorithms. To achieve a tradeoff between the fairness of a Round Robin like algorithm and the throughput of an opportunistic scheduler, we propose a multistream Proportional Fair scheduling algorithm. We also introduce some heuristic simplifications to reach a suboptimal solution within the TTI. The multistream Proportional Fair scheduling algorithm we propose has been implemented in MATLAB, and has been tested in the scenario of HSDPA beamforming transmissions. The inter beam interference has been modelled by a spatial orthogonality factor. By sequencing beams oriented in sufficiently different directions, we can improve the throughput by a factor (nT1)/log2(nT1) the throughput of a single beam 4x4 MIMO channel.
|journal||the 6th Management Committee Meeting of COST2100 - Lille (France), 6-8 Octobre 2008|
|Etat de la publication||Publié - 2008|