Performance of Different Weighting Techniques with DIA MADM Method in Heterogeneous Wireless Networks

The paper investigates the optimum pairing of attribute weighting techniques used with Distance to Ideal Alternative (DIA) selection in Multi-Attribute Decision Making (MADM) applied to wireless network vertical handover. The performance of the DIA algorithm is investigated with Analytical Hierarchy Process (AHP), Fuzzy AHP, Analytic Network Process (ANP), Random Weighting (RW) and Least Squared weighting (LSW) techniques for different user service classes including background, conversational, interactive, streaming traffic. It is found that the use of different weighting techniques for every traffic class results in widely varying performance figures. Comparative numerical results are given to quantify the system performance in terms of best network selection and ranking abnormalities. The paper findings provide guidelines for selecting weighting techniques for each traffic class in order to achieve ideal network selection with minimal ranking abnormality.