Estudo do compromisso entre eficiência espectral e justiça na alocação de recursos de rádio

Abstract

This work studies the trade-off between spectral efficiency and fairness considering the alocation of radio resources in a multi-user system which considers quality of experience constraints for each user and service. The focus herein is to reduce the non-uniformity among the throughput values achieved by the users, which originate from the opportunistic resource distribution performed by the resource allocation algorithms. Thus, it is proposed herein the introduction of a penalty term in the objective function of the resource allocation optimization problem as to provide a better trade-off between fairness (in terms of users’ throughput) and the total system capacity. The obtained results were analyzed in terms of Jain’s index of fairness, where different variations of several relevant parameters which impact the users’ satisfaction have been evaluated. The results have shown that the proposed approach provided a better trade-off between fairness and capacity reaching Jain’s index of fairness values of up to 0.9 with rates close to those obtained in the case without the penalty in the objective function.