With the advent of higher bandwidth internet services, video over IP is becoming more common with the video transfer mechanisms currently being standardized by the Open IPTV Forum. IPTV in the home is starting to dominate the market and is set to replace the way we watch our media content from watching it when broadcast to watching it on demand. Existing home gateway solutions or access points can allocate bandwidth to users depending on traffic priority or on a first come first served basis. This however might not be an optimal choice of allocation since it does not take into account user privileges and user perceived quality. This paper proposes a Cognitive Resource Manager, which takes into account these factors for managing the resources in the home. The paper suggests methods for learning user preferences by employing simple machine learning techniques and defines a user satisfaction utility, then optimizes it, choosing the optimal configuration for resource allocation. The paper also proposes a more lightweight heuristic approach which approximates the optimal configuration for resource allocation at a lower computational cost. This method of resource allocation compares favorably to a number of alternative strategies with respect to fairness.
|Title of host publication||IEEE/ETRI International Conference on Information and Communication Technology Convergence|
|Subtitle of host publication||ICTC 2010|
|Place of Publication||Jeju, Korea|
|Publication status||Published - 18 Nov 2010|
- Cognitive Radio