Hybrid Algorithm for Selecting Multimedia Network Connections in Heterogeneous Networks

Abstract

The heterogeneity of mobile and wireless networks exposes mobile users to different access network technologies. In the past, many paradigms have been introduced, such as Always Best Connected (ABC) and Always Best Network Connection (ABNC), to meet user’s preferences. However, they fail to consider multimedia services consisting of video and data, besides voice, for both source and destination. This paper presents a new model called Always Best Multiple Network Connection (ABMNC) to support multimedia services. ABMNC is first formulated as a Multiple Attribute Decision Making (MADM) problem with an embedded utility-based Multiple Knapsack Problem (MKP). In order to reduce computation complexity, the MADM hierarchy of ABMNC is decomposed into a number of iterated sub-MADM hierarchies, and a hybrid Analytic Hierarchy Process (AHP) and a Simple Additive Weighting (SAW) scheme are used to solve them. A novel Heuristic Rate Allocation Algorithm (HRAA) is then presented to reduce the computation complexity of the assignment and rate allocation. Moreover, a comprehensive Heuristic Path Selection Algorithm (HPSA) is proposed to efficiently resolve the ABMNC hierarchy. Finally, computer simulation is performed to study ABMNC, and the results show that our approach, most of the time, chooses the optimal network connections.