Dynamic pricing of wireless network resources in a competitive provider setting

Abstract

The ongoing convergence of cell-based wireless systems, from proprietary standards for data transmission toward a unifying approach based on the Internet Protocol (IP), as well as the increasing ability of wireless computer networks to support Quality-of-Service, opens up a tremendous potential for new business models and is expected to significantly alter the way mobile and wireless services are provided. An important field of study is about understanding the implications on the competitive structure of the wireless market. With the technical barriers between di erent technologies more and more diminishing, it becomes possible for customers to select network access on-demand instead of entering long-term contracts with a single wireless network provider based on factors such as connection speed, level of mobility and prices. In such a setting, wireless network providers, but also the regulatory bodies designing the principle market rules, need to learn about new ways of allocating resources among customers to reach their objectives. With resources being available on-demand, pricing becomes an important factor for controlling demand and distributing resources among users. Instead of being fixed over longer periods, pricing can be used for signalling the congestion levels in di erent wireless networks and to let users communicate their willingness-to-pay to the market. Besides presenting a classification framework for pricing in wireless networks and providing a comprehensive review of the relevant literature, the main contributions of this thesis are on the development of behavioural strategies, which di erent entities can follow in the described scenarios. In the first scenario we regard wireless resources as a public good, which should be efficiently allocated among active customers. By letting each provider use a second price auction, with each customer being able to bundle resources from multiple providers, we develop a bidding strategy, which allows customers to distribute their demand among providers. The bidding strategy resembles truth-telling in such a market and is shown to be the best strategy a user can follow to fulfill his demand. We examine the properties of the system analytically and present the results of an extensive simulation study. In the second scenario we model a free market for wireless resources on which prices are determined at the time of request. With customers having more than one option to obtain resources, the task of a provider becomes setting prices so that it maximises revenue under partial or full competition. We discuss the constrained maximisation problem and analyse the situation of a two-provider setting as a game. Simulation is used to analyse the game under uncertainty and to understand if a steady-state situation arises. While the results make use of multiple simplifications to allow us to understand the main model variables, the findings can potentially be of interest for providers and market authorities to develop new ways for resource allocation using dynamic pricing as a tool for reaching their design objectives.