Squeezing Out Capacity with Joint Power-Control and Channel Assignment

In this paper, using simulations, we compare three sets of channel assignment algorithms for joint power control and channel assignment with varying complexity. The channel assignment of M users to N channels in K cells is completed in several iterations of application of the algorithms. In each round, up to N users are assigned to the N channels. The algorithms we study fall into three subsets depending on (i) whether the power control is based on the users path gain only or on the path gains of all users sharing the same channel; and (ii) whether in each round, the N channels assignments are done sequentially or simultaneously. We assume that the path gain matrix, the global channel assignment and power control information is available at a central location. Our aim is to quantify the differences in the system capacities achieved by these different joint power control and channel assignment algorithms. Our results may be used to trade-off the complexity of the algorithms with the capacity gains that are achievable.