TITLE: Design Of Survivable Wireless Access Networks

WHEN: Friday, September 26, 2003 Time: 1:30 -4:00 PM

WHERE: Room 503, IS Building

WHO: Chalermpol Charnsripinyo

Dissertation Committee:

Dr. David Tipper (Committee Chair)
Dr. Richard Thompson
Dr. Prashant Krishnamurthy
Dr. Joseph Kabara
Dr. Teresa Dahlberg, University of North Carolina at Charlotte

ABSTRACT: Survivability of wireless access networks has become a critical issue due to the rapidly increasing dependence on wireless mobile services. However, little emphasis has been placed on understanding and improving the survivability of wireless access networks. In this dissertation, survivability issues including a multilayer survivability framework and survivability analysis of wireless access networks are presented in order to facilitate the development of effective techniques for building survivable wireless access networks.

This dissertation proposes new techniques for survivable wireless access network design. The goal is to provide an effective solution for wireless access network design that satisfies traffic and survivability requirements while minimizing the network design cost in the wireless backhaul network. Three different approaches to creating a survivable network topology are considered namely: (1) a mesh-based restorable network, (2) a multi-ring network, and (3) a hybrid mesh-ring network. The network design problems are formulated as mathematical optimization models. A novel aspect of the proposed work is the incorporation of the effect of user mobility after a failure into the network design model. Numerical results show that, among the three different network design approaches considered, a mesh-based restorable network design is the most cost-efficient. A two-phase design methodology is developed to solve the mesh-based restorable network design problem. In the first phase, an initial network design problem is considered as the minimum-cost network design to satisfy traffic requirements. In the second-phase, an incremental network design problem is considered where the minimum-cost network topology from phase one is augmented to satisfy survivability requirements. The proposed second phase model can be easily applied to existing wireless access networks.

This dissertation also proposes an efficient heuristic algorithm based on a minimum-cost routing technique to solve the mesh-based restorable network design problem. The proposed heuristic can efficiently find near-optimal solutions compared to the solutions obtained from a standard optimization technique for small problem-size networks. More importantly the heuristic is shown to scale to solve large problem-size networks within a reasonable computational time. Lastly, the proposed two-phase mesh network design model and heuristic algorithm are extended to the design of emerging third-generation wireless access networks.