Resource allocation model based on Particle Swarm Optimization for OFDMA macro-femtocell networks

In this paper, we address the problem of resource allocation under high user density environment in OFDMA macro-femotcell networks. In the literature, several solutions have been proposed based on spectrum partitioning or sharing. However, they failed to enhance the overall system throughput due to: (1) resources underutilization because of the unchanged macrocell power distribution over the dedicated allocated bandwidth, (2) lack of appropriate access mechanisms that allow femtocells to grant access to nearby public users without depriving their own subscribers, and (3) lack of adaptive power control that finds a trade-off between overall system performance if femtocell can grant access to nearby public users. Therefore, we propose a heuristic spectrum partitioning approach that avoids the interference by means of orthogonal subcarrier assignment among the two tiers and neighbor femtocells and allows spectrum reuse in femtocells if they are far from each other. Moreover, power adaptation is incorporated to enhance mobile users satisfaction over the active subcarriers in each base station. The proposed solution, based on Particle Swarm Optimization, maximizes entire network throughput and is able to determine: the best serving base station, power and bandwidth for each mobile user taking into account its location, demand and FC proximity. Simulations are conducted to show the performance of the proposed solution that is also compared with a linear programming approach and a modified Weighted Water Filling.