Abstract
With the recent advances in IP television network and TV devices' technologies demand for new services for TV viewers is increasing. One of such services is providing TV contents at convenient to subscribers time and locations. A considerable amount of network bandwidth is required to deliver real-time contents to multiple subscribers due to the large volume of unicast traffic over an IPTV network. To address some of these challenges, we propose a new model for content caching that takes advantage of the cooperation among distributed caches containing most recommended contents in Digital Subscriber Line Access Multiplexers (DSLAM), Central Office (CO), and Intermediate Office (IO) caches. The objective of this study is to explore opportunities of using big data analytics in IPTV industry system to significantly improve the quality of services, competitiveness, and enhance capability to offer contents to viewers. Content providers need to make sure their viewers get the best experience possible. Our model considers the viewer's content preferences based on the favorite programs as well as the event-based analytics using recommendation algorithms. It employs agent technology to move all or part of the cache allocated to a client TV device on a corresponding DSLAM to the TV client or, if desired, content has not found on DSLAM, from upper level (CO and IO) to the DSLAM of the TV client.