Context Learning and Configuration Selection for Cognitive Device Management

According to the B3G concept, diverse radio access technologies (RATs) can converge into one composite radio environment, where the user is "always best connected" and can move between the various technologies seamlessly. Cognitive wireless networks [1],[2] have emerged as a complementary concept to B3G networks. In general, cognitive systems establish their behaviour, in a reactive or proactive manner, based on external triggers (environment), as well as their goals, principles (policies, rules), experience and knowledge. This requires, among other things, advanced terminal management functionality for dynamically selecting and adapting to the most appropriate technology, in a seamless, transparent and robust manner. This paper emphasizes on the management of user equipment and discusses on how to further progress the relevant work. The work of this paper complements legacy methods for network selection by presenting cognitive mechanisms and an associated knowledge-based network selection scheme. In particular, a specific learning scheme that relies on the use of Bayesian statistics is proposed. In the direction of assessing the performance of the developed knowledge-based selection scheme a software implementation has been realised. This software implementation also includes a legacy selection scheme thus enabling the comparison with the knowledge-based selection. Corresponding results derived from the software implementation are presented.