We consider distributed dynamic slot and power allocation for the downlink of a TD/SDMA broadband wireless packet network with multiple access ports and adaptive antennas. The still open issue for packet multicell SDMA is how to manage the intercell interference, which is very difficult to predict in an uncoordinated environment, due to packet access and downlink beamforming. For this reason, doing distributed allocation efficiently as well as improving the performance by means of power control results in a very hard task. We propose a greedy SDMA algorithm exploiting the power shaping technique, which is based on a static preallocation of the transmit power to each slot of the frame. This permits to obtain a partial predictability of intercell interference, allowing different levels of estimated intercell interference and available power for each slot. We show that our greedy SDMA algorithm with power shaping increases system capacity with respect to the same algorithm without power shaping and reduces the performance gap with respect to a greedy centralized strategy, thus limiting the need of coordination among cells. Both centralized and distributed algorithms are compared, as reference, to the baseline case of random allocation, previously proposed for packet access.
Distributed dynamic resource allocation with power shaping for multicell SDMA packet access networks
TRALLI, Velio;
2004
Abstract
We consider distributed dynamic slot and power allocation for the downlink of a TD/SDMA broadband wireless packet network with multiple access ports and adaptive antennas. The still open issue for packet multicell SDMA is how to manage the intercell interference, which is very difficult to predict in an uncoordinated environment, due to packet access and downlink beamforming. For this reason, doing distributed allocation efficiently as well as improving the performance by means of power control results in a very hard task. We propose a greedy SDMA algorithm exploiting the power shaping technique, which is based on a static preallocation of the transmit power to each slot of the frame. This permits to obtain a partial predictability of intercell interference, allowing different levels of estimated intercell interference and available power for each slot. We show that our greedy SDMA algorithm with power shaping increases system capacity with respect to the same algorithm without power shaping and reduces the performance gap with respect to a greedy centralized strategy, thus limiting the need of coordination among cells. Both centralized and distributed algorithms are compared, as reference, to the baseline case of random allocation, previously proposed for packet access.I documenti in SFERA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.