An Evolutionary Game-Theoretic Approach for Base Station Allocation in Wireless Femtocell Networks

The deployment of femtocell is as a potential solution to improving the indoor coverage and capacity of a cellular network. However, providing quality of service (QoS) is one of the most significant challenges in wireless femtocell networks. Cellular users with different QoS requirements always look for a serving base station (BS) to achieve their desired QoS. Therefore, intelligent BS allocation plays a crucial role in guaranteeing the QoS. In this paper, we propose a game-theoretic BS allocation approach with a QoS guarantee for downlink femtocell networks. We model the BS allocation problem as an evolutionary game in which users with bounded rationality learn from the environment and make BS selection decisions with minimum exchange of information. In the propounded model, not only do we consider the users' QoS, but we also take account of macro users' activity as an interferer to reduce the cross-tier interference. Unlike the previous studies, we calculate the demand rejection probability for each user associated with a BS to evaluate statistical QoS guarantees. A distributed learning-based algorithm is developed to show the existence and fairness of the solution to our proposed model and to demonstrate the convergence to the evolutionary equilibrium as the solution of the game. Simulation results verify the effectiveness, performance improvement, and good convergence of the proposed approach.