Personalised subscription pricing for optimised wireless mesh network deployment

  • Authors:
  • S. M. Allen;R. M. Whitaker;S. Hurley

  • Affiliations:
  • School of Computer Science, Cardiff University, Queens' Buildings, 5 The Parade, Cardiff, CF24 3AA, United Kingdom;School of Computer Science, Cardiff University, Queens' Buildings, 5 The Parade, Cardiff, CF24 3AA, United Kingdom;School of Computer Science, Cardiff University, Queens' Buildings, 5 The Parade, Cardiff, CF24 3AA, United Kingdom

  • Venue:
  • Computer Networks: The International Journal of Computer and Telecommunications Networking
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

Wireless mesh networks (WMN) represent semi-infrastructured ad hoc networks, where nodes act as both access points and routers. Wireless links between nodes are used to relay communication to a gateway node or point-of-presence (POP) for a wired or fibre network. Our interest is in mesh network deployment for a service provider (typically an ISP) who has a fixed location(s) for a POP(s). Service coverage requires that new subscribers are able to route to a POP via existing subscribers. Providing service coverage in a new region is problematic since a node is covered only if at least one of its neighbours is active in the network. Consequently an insufficient number of active nodes will lead to disconnected components. Seed nodes, owned by the network, are one mechanism for enhancing coverage, but at an additional upfront cost. In this paper we consider the influence of pricing strategies on the expected coverage and revenue. Assuming ideal economic behaviour and demand for subscription based on price, we introduce a framework that seeks to assign personalised subscription prices to optimise the expected network formed in terms of both coverage and revenue. The proposed method is benchmarked with other alternative strategies and a range of test problems are explored, showing the extent to which financial and coverage gains can be achieved if personalised pricing can be adopted. We further show how seed node and POP selection can be incorporated in the same optimization framework, allowing the trade-off between pricing and infrastructure costs to be investigated.