Genetic algorithm-based adaptive channel allocation for best-effort traffic users in microcellular systems

초록

Previously proposed adaptive channel allocation schemes in cellular systems have mainly focused on the case in which a dedicated channel is exclusively allocated to each user. Thus, it is important to assign the infrequently used channels in adjacent cells among all available channels to reduce the probabilities of call drop and call block. Even when best-effort traffic users are supported by sharing multiple channels through time-domain scheduling, a similar approach (i.e., the conventional scheme) has utilized channels experiencing the lowest interference at each basestation. However, it is unlikely that the best-effort users are fairly supported using the channels chosen by the conventional scheme regardless of location because the interference conditions on these channels tend to be quite different for users far from its serving basestation. In other words, a channel experiencing low interference at the serving basestation does not guarantee adequate performance for a user who is far away. In this paper, we propose an adaptive channel allocation scheme in which the best-effort users are fairly supported regardless of distance to the serving basestation. Specifically, using genetic algorithm, the channel set that maximizes the sum of the far users' throughputs is found while allowing a pre-specified loss in the sum of all users' throughputs compared with that of the conventional scheme. Using computer simulations, it is shown that the throughput performance of the 10th percentile throughput users improves up to 38% in some cases, when the overall throughput performance loss was allowed to be less than 5% compared with that of the conventional scheme. In addition, only some of the link-gains can be measured in practice, because it is very difficult for a user to measure the link-gain of a distant basestation. Even when only some link-gains can be measured, performance degradation is negligible. © 2013 ICIC International.

키워드