Application of MADM Algorithms to Network Selection

Abstract: In heterogeneous wireless networks, vertical handoff plays a crucial role in providing seamless continuity to a multimode terminal user. Network selection is the key element in vertical handoff process. Several strategies have been proposed in literature to address the problem of network selection and multiple attribute decision making (MADM) methods have evolved as one of the most promising solutions. This paper compares the performance of three MADM methods, namely SAW (Simple Additive Weighting), WPM(Weighted Product Method) and PROMETHEE (Preference Ranking Organization Method for Enrichment Evaluations) in network selection. The AHP (Analytic Hierarchy Process) method is used to assign weights to the criteria. Simulation results show that PROMETHEE provides more accurate network selection for conversational, interactive and background traffic classes. However the performance of SAW, WPM and PROMETHEE is found to be similar and inappropriate for streaming traffic class.

Keywords: Network Selection, SA, WPM, PROMETHEE, AHP.