User satisfaction models and scheduling algorithms for packet-switched services in UMTS
Résumé
In this paper, we propose a user satisfaction model for packet-switched services in next generation wireless networks for different radio resource allocation (RRA) strategies. Here, user satisfaction is seen as the joint effect of intrinsic service value and user perceived QoS which is seen as an increasing and concave function of network performance. The radio resource allocation process has to balance between power constraints and mean user satisfaction maximization. This optimization problem can be solved by the Lagrangian method and leads to a new scheduling algorithm. Then, we present other typical satisfaction functions for which the Lagrangian method also works. This leads to known algorithms such as max C/I scheduler or proportional fairness. By performing end-to-end dynamic simulations (TCP, RLC/MAC, WCDMA layer), we compare performance results in terms of user satisfaction and throughput distribution. The results show that a scheduling strategy always makes a trade-off between fairness and efficiency. We also show that it is possible to achieve service differentiation for packet-switched services by adjusting intrinsic service values.