Opportunistic beamforming using dumb antennas
IEEE Transactions on Information Theory
| Hi-index | 0.00 |
Future wireless communication systems are required to support extremely high speed packet data transmission. Multiple antennas and multi-user packet scheduling are two promising techniques to achieve this target with limited available spectrum resource in PHY layer and MAC layer, respectively. One interesting synergy of these two is the opportunistic beamforming method proposed in [1]. Unlike traditional deterministic beamforming method where the beams are generated pointing to a designated user, beams are generated randomly in opportunistic beamforming method. By reporting the resultant channel gain, one user who opportunistically fall into the main lobe will be scheduled. However, the original method only generates one random beam so that large part of spatial resource is wasted. Later in [2], the author proposed to generate multiple random beams so that opportunistic spatial division multiple access (SDMA) is possible. However, there are always nT number of scheduled users which is not necessary optimal due to variation of channel condition. On the other hand, the adaptive method proposed in [3] can properly adjust the number of scheduled users, however, with large amount of feedback information. This paper propose a new complex reduced adaptive method by utilization of time-correlation of resultant fading channel. In particular, the number of scheduled users will be tuned by a close-loop control procedure between the base station and mobile users over different scheduling intervals. Simulation results show that proposed method can perform quite close to optimal method with tremendous reduction of feedback information.