Throughput and QoS pricing in wireless communication

  • Authors:
  • Andrey Garnaev;Yezekael Hayel;Konstantin Avrachenkov;Eitan Altman

  • Affiliations:
  • St. Petersburg State University, St. Petersburg, Russia;University of Avignon France, Avignon, France;INRIA Sophia Antipolis France, Sophia Antipolis, France;INRIA Sophia Antipolis France, Sophia Antipolis, France

  • Venue:
  • Proceedings of the 5th International ICST Conference on Performance Evaluation Methodologies and Tools
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper we investigate a bi-level economic model of wireless communications. We take into account the user's demand and also the QoS offered by the network. We determine with our model when a provider can make more profit by investing into new technology. We consider a Stackelberg game approach where the provider decides the price per rate. In fact, users have to pay for the throughput they have contracted with the service provider in their SLA. Then, pricing for throughput is more realistic then pricing for power as it is suggested in several papers in the literature. Our approach is general as we have important results for general user's utility function with only natural assumptions on them. We also study next an incomplete information case in which the provider has no perfect information about the user demands in network and its quality. To deal with it we apply a Bayesian approach. We demonstrate that the expectation on the increasing user demand can induce jumping reduction of the optimal tariff meanwhile impairs expectation about increasing in user activity can lead to jumping increasing in the optimal tariff. This jumping nature of tariff tells about importance of marketing research for provider to correctly estimate user's demand in his intention of using network and how he personally values the service. Finally, we consider the multiuser case and obtain a closed form expression of the Stackelberg equilibrium in that case. In particular we show that the provider cannot essentially increase its profit by simple increasing number of users.