Abstract
Network selection in Heterogeneous Wireless Networks (HWN) aims to select the best network for a variety of communication tasks at any time and anywhere. Due to its importance and necessity, it has been studied using various mathematical models including game theory, multiple attribute decision making (MADM), Markov chain, and fuzzy logic. Since these models have different features and functions to produce different results, it has been suggested to combine these models in order to harness the benefit of individual model. However, it remains a challenge to decide when and how to combine these models or systems. In this thesis, we propose a new approach to study the network selection problem in HWN using combinatorial fusion. More specifically, we investigate: (a) vertical handoff decision to fuse three metrics: received signal strength (RSS), data rate, and network latency using fuzzy logic and combinatorial fusion; and (b) load balancing using combinatorial fusion on three metrics: RSS, accumulated message queue length, and channel utilization. Experimental results demonstrated that our method can make network selection much simpler and more effective. Our work provides a novel way to fuse these metrics for network selection. It is also the first method to integrate fuzzy logic and combinatorial fusion in solving the network selection problem in heterogeneous wireless network.