NetANPI: A network selection mechanism for LTE traffic offloading based on the Analytic Network Process

Abstract

Traffic offloading in Long Term Evolution (LTE) networks is a key process in cellular networks so a mobile device chooses one of the available femtocells. In the last few years, several network selection methods have been proposed based on multiple attribute decision making (MADM) techniques for traffic offloading in LTE. The Analytic Network Process (ANP) is an MADM method that has been barely studied for network selection in cellular networks; ANP configures the decision making problem as a network of attributes in order to derive priority scales of individual judgments. We propose in this paper NetANPI (Network selection ANP-based mechanism with Ideal network comparison), which is an ANP decision making mechanism for traffic offloading in LTE. We show its effectiveness by contrasting it with MADM methods proposed in the literature as AHP, ELECTRE, TOPSIS, GRA, and SAW. We also introduce a mechanism to rank the MADM algorithm using utility theory. Our results show that NetANPI outperforms the most used MADM mechanism for the interactive and conversational traffic classes.