| Summary: | The phenomenal emergence of several heterogeneous wireless networks and technologies has allowed users to have access IP services anywhere at anytime from any network with whatever terminal they use. This computing platform has also been driven by the rapid evolution of mobile devices that are equipped with multiple network interfaces and the development of IP based applications. One of the challenging tasks with this Next-Generation Networks (NGN) computing platform is service continuity when users roam around different wireless networks e.g. Wi-Fi, Bluetooth and Cellular networks. This challenge is elevated more when dealing with applications that distribute time continuous data with stringent Quality of Service (QoS) requirements. One of the adaptation methods to ensure service continuity by minimizing flow interruptions when users are mobile is session handoff.
The main contribution of the thesis is to present a handoff support system which implements a handoff decision engine using a Multi-Criteria Decision Making (MCDM) method based on a Bayesian Belief Network (BBN) and a handoff execution procedure based on buffering and doublecasting techniques. The handoff support system is built around the following features: 1) It utilises a proxy-based middleware architecture, 2) It uses a BBN based MCDM for handoff decision, 3) It is able to represent the full context information which represents the execution environment, 4) It is able to perform decision making under both certainty and uncertainty, 5) It is able to decide correctly on the target network under dynamic context, 6) It performs decision making in the midst of conflicting, interdependent and constraint criteria, and 7) It uses a profile-based handoff decision to offer personalisation to users.
The experimental results showed that when compared with Analytical Hierarchy Process (AHP), the handoff decision method based on BBN performs better on: 1) Modelling of the handoff decision problem and the full representation of the context information, 2) Decision making under uncertainty, 3) Modelling of constraints and interdependent criteria and 4) Support for user preferences. When evaluating the handoff execution, further results revealed that the underlying handoff management strategies provide service continuity by minimising handoff latency and packet losses.
|
|---|
