Cooperative Multi-Cell Power Control under Imperfect Channel Knowledge: A Comparative Performance Study

Radio resource management is important in enhancing spectrum utilization efficiency in broadband wireless services. Multi-cell binary power control (BPC) in full-reuse networks is a promising candidate in system optimization for this purpose. While the BPC scheme is inherently simple, the sum rate maximization by cooperative BPC across cells/links is almost as good as the best achievable non-binary solution. For practical use, the issue of imperfect channel knowledge should however also be considered. In this paper, we conduct an investigation and analysis of the performance of BPC and compare it to the best known non-binary optimization, as well as to full-power transmission. Performance under two different link gain measurement error models is reported. Clearly, with perfect channel knowledge, BPC usually has system capacity lower than that from the optimized non-binary allocation, although the loss is marginal. However, more interestingly, when measurement errors occur, BPC achieves a better and more robust performance. Simulation results show that it is less sensitive to the negative impacts of the link gain measurement errors. When looking at the simple full power scheme, comparisons show that although it is severely suboptimal when perfect channel state knowledge or moderate channel state error levels are considered, it can actually outperform BPC and also GP-based optimization in some scenarios with severely erroneous channel knowledge.