Dynamic implementation of network selection policies

摘要

The issue of efficiently selecting the best available wireless network by a mobile device is an area of active research. Many researchers have proposed several strategies and policies for selecting the appropriate network. These policies differ among each other due to various reasons. The state and load of networks, number and types of wireless networks considered, and preferences of users are some of the reasons that attribute to differences in those proposed policies. As a result, the applicability and efficiency of those policies depend upon certain situations and circumstances. We propose, in this paper, a new concept that will help in utilizing these various policies in a manner that will give better results in a longer run and in various kinds of situations. We have a pool of policies available, and our proposed method will select the policy that is most appropriate according to the current state of user. We modeled this problem as a Markov Decision Process. In the end we provide some heuristics to support our work.