부산대학교 도서관

Television (TV) consumers adore interactivity and personalization. Internet Protocol television (IPTV) systems help satisfy this desire by providing different services including video-on-demand (VoD), network personal video recording (NPVR), and time shift TV (TSTV). However, these services place a heavy traffic load on the network by requiring a unicast video stream for each active customer. This paper discusses means to reduce the network traffic load by introducing storage deep in the network—thus avoiding the need for unicast traffic support over the metro or core network—and algorithms to place content intelligently on those storage devices. An algorithm for intelligent redistribution is described that endeavors to place the most appropriate content close to consumers. The paper discusses further how this new paradigm of network storage can help improve quality of experience for the IPTV consumer and can assist in the deployment of Internet video and sharing of user-generated content (UGC).