Application of Fuzzy AHP and ELECTRE to Network Selection

Abstract. In a heterogeneous wireless network environment services are ubi-quitously delivered over multiple wireless access technologies. Ranking of the alternatives and selection of the most efficient and suitable access network to meet the QoS requirements of a specific service, as these are defined by the us-er, constitutes thus an important issue. Decisions on which network to connect to are however difficult to be reached, since multiple factors of different relative importance have to be taken into consideration. This paper addresses this diffi-culty by adopting Multi Attribute Decision Making (MADM) methods. Fuzzy AHP, a MADM method, is initially applied to determine the weights of certain Quality of Service indicators that act as the criteria impacting the decision process. The fuzzy extension of the method, and consequently the use of fuzzy numbers, is adopted in order to incorporate the existence of fuzziness as a result of subjective evaluations. Afterwards, ELECTRE, a ranking MADM method, is applied to rank the alternatives, in this case wireless networks, based on their overall performance