A Greedy Heuristic Algorithm for Context-Aware User Association and Resource Allocation in Heterogeneous Wireless Networks

Abstract : Wireless heterogeneous networks are usually characterized by the integration of different types of radio access technologies (RATs) to enhance system capacity and meet user requirements. In this context, we formulate a user association and downlink resource allocation optimization problem to maximize the overall user-centric profit in the system. The context-awareness is based on the user preferences, the data rate requested by each user equipment (UE), and the RAT characteristics and constraints. The user preference is based on a normalized weighted profit function that considers both the received signal quality and the power consumption at UEs. To approximate the formulated optimization problem, a greedy heuristic algorithm with polynomial-time complexity is proposed. It is shown through persuasive simulations that the proposed heuristic algorithm, when compared to the trivial profit-function-based solution, enhances the average user satisfaction in the system and lowers the percentage of blocked data rate. In fact, the performance of the proposed heuristic algorithm comes close to the optimal solution while requesting a lower number of handovers (HOs).