Learning in user-centric IPTV services selection in heterogeneous wireless networks

It is envisioned that in IPTV services will be run on the top a heterogeneous mix of network infrastructure. The existence of multi-mode terminals enable users get associated to the best available networks according to user preferences and application specific requirements. In this paper we study the user-centric service selection for IPTV services. We propose the user satisfaction function for video services and validate this function against the objective measurement results. The users select the service based on his satisfaction function. We model the service selection problem using dynamic games. We also introduce novel scheme called cost of learning that incorporates the cost to switch to an alternate IPTV service provider. Using the Evolutionary game dynamics, we study the convergence and stability properties in user-centric IPTV service selection problem. We propose a trusted third party based architecture solution to realize the user-centric network selection. For the proof of concept simulations were run in OPNET and Mathematica.