Towards an Efficient Network Selection Technique Based on Differentiated Weight of Access Interface

In this work, the authors have proposed a new technique for network selection decision. This technique combines two multi attribute decision making (MADM) methods. The analytic network process (ANP) method to find the differentiate weights of available networks by considering each criterion and the grey relational analysis (GRA) method to rank the alternatives. To show the effectiveness of our technique we have presented the simulation results of four traffic classes namely background, conversational, interactive and streaming. DOI: 10.4018/jbdcn.2012100103 International Journal of Business Data Communications and Networking, 8(4), 32-48, October-December 2012 33 Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited. The vertical handover process can be divided into three parts namely: handover initiation, network selection and handover execution. The present work concentrates on the second part which is considered the principle key of the vertical handover. For this reason, we propose a new technique for network selection decision which allows users to select among the available networks, the most suitable network in terms of QoS. The network selection decision depends on the following multiple criteria: • From Terminal Side: Battery, velocity, etc. • From Service Side: QoS level, security level, etc. • From Network Side: Provider’s profile, current QoS parameters, etc. • From User Side: Users preferences, perceived QoS, etc. Many schemes and decision algorithms have been proposed and developed in the literature to optimize the network selection decision. We could classify them into four groups such as Multi Attribute Decision Making (MADM) methods, genetic algorithms, fuzzy logic and utility functions. In Zhang (2004) the authors have presented a new strategy to solve the vertical handover decision problem by using the fuzzy Multi-attribute decision making methods (Fuzzy MADM). In Attaullah et al. (2008) the authors have suggested an intelligent approach for vertical handover based on fuzzy logic. In Fu et al. (2010) and Yafang, Huimin, and Jinyan (2010) the authors have proposed to combine two MAMD methods namely Analytic Hierarchy Process (AHP) method and Gray Relation Analysis (GRA) method. The AHP method is used to determine weights for each criterion and the GRA method is applied to rank the alternatives. In Gyekye et al. (2008) the authors have proposed a new network selection algorithm which based on AHP and genetic algorithms. The first algorithm is used to weigh different criteria and the second one is used to optimize the network function in order to select the optimal access network. Moreover, in Yan et al. (2008) and Pervaiz (2010) the authors combined AHP method and function utility. The AHP method is introduced to calculate the relative weights of criteria and the utility function is used to rank each access network. Finally, in Sgora et al. (2010) and Lahby, Leghris, and Adib (2011) the authors have proposed to combine the AHP and TOPSIS methods for the network selection problem. The AHP method is used to weigh different criteria while TOPSIS method is applied to rank the available networks. Although, AHP method is often used to weigh criteria, this method still presents two different weaknesses: 1. The First Problem: The hierarchic architecture requires independence among all those criteria that are in the same hierarchy level. 2. The Second Problem: In the majority of situations, the index consistency CI > 10% which necessitates to re-establish the pairwise comparison matrix. In addition, all network selection algorithms assign the same vector weight independently of the available networks. In other words, the user preference relative to each criterion is considered to be the same for each access network. On the other hand GRA method suffers from ranking abnormality (Husza et al., 2010). The ranking abnormality means that the ranking of candidate networks changes when low ranking alternative are removed from the candidate list, which can make the selection problem inefficient. The major factor causing the ranking abnormality is the weighting algorithm (Lahby, Leghris, & Adib, 2012) used to weigh different criteria. To deal with the above problems, this paper proposes an efficient network selection technique. This technique is based on two MADM methods such as ANP and GRA. In the classical network selection algorithms, which are based either on AHP method or ANP method, the criteria have the same relative importance 15 more pages are available in the full version of this document, which may be purchased using the "Add to Cart" button on the product's webpage: www.igi-global.com/article/towards-efficient-networkselection-technique/75127?camid=4v1 This title is available in InfoSci-Journals, InfoSci-Journal Disciplines Communications and Social Science. Recommend this product to your librarian: www.igi-global.com/e-resources/libraryrecommendation/?id=2