QoE Optimized Resource Allocation in Multiuser OFDM Systems

The resource allocation in multiuser Orthogonal Frequency Division Multiplexing (OFDM) systems is addressed, which aims to maximize the user-perceived quality of services instead of the conventional objective metrics. By introducing the previous application-centric Mean Opinion Score (MOS) assessment model, the problem is formulated as a bounded optimization problem, which is non-concave and difficult to develop a closed-form solution. A heuristic algorithm, which couples bound relaxation technique with further fix and drop operations is then proposed for the final solution. Simulation results indicate that our proposal can improve the overall perceived quality from users' perspective as well as the fairness between users, in comparison with the classical throughput-oriented algorithm.