NETWORK SELECTION BASED ON USER PREFERENCES BY USING GAME THEORY

Abstract

In the recent years wireless communication systems and networks were spreading all over globe to ensure to reach "Always Best Connected" (ABC) vision. Different wireless networks are trying converge and forming heterogeneous environment. In this environment, the users will carry multi-mode mobile devices to access one or more wireless networks. All the networks differ in terms of technology, energy usage, coverage area, cost and bandwidth etc. in this scenario, user need to select a best suitable network for his requirements. But to decide to which network to select based on dynamic and static multi parameters is difficult. For specific service what parameters has to consider and how these attributes will model requirements of application and user need is important. In this thesis multi attribute decision making methods comparison has discussed and game theory components are mapped into network selection situation. In modelling the problem situation, Simple Additive Weighting (SAW) method is used as scoring function to ranking the networks. To evaluate user preferences over attributes Analytical 1-lierarchy Process (AHP) is used in algorithm which affects network selection decision. Utility functions are developed for quantification of user satisfaction with respect to each parameters. In presented reverse auction game, players, payoffs and strategy space components discussed. The objective is to maximize user satisfaction and increases revenues of networks. Simulation results shows that traditional MAI)M technique i.e.) SAW is an ordinal in nature. So for criteria negotiation, along with SAW method auction game is used to model the interactions between user and networks. Results shows that combination of SAW and non-cooperative game achieves better selection compared to SAW method.