Description
Consumer demand for watching videos over the Internet is increasing faster than video providers' ability to efficiently deliver the multimedia experience. Peer-to-peer (P2P) Video-on-Demand (VOD) applications are emerging methodologies for delivering this experience. A problem with a P2P solution is efficiently storing replicated content among the peers. This research proposes a novel approach for replicating video content in a P2P environment based on learning and predicting users' request habits of their video preferences by employing a Complex Adaptive System (CAS). This is accomplished by using the CAS to emerge intelligent decision making. Referencing available metadata for each video, the algorithm utilizes an Artificial Immune System (AIS) to recognize and adapt to user requests over time, allowing the system to evolve to user demand over time. We introduce the Video-on-Demand Complex Adaptive system (VODCA) framework to discover and tune for optimal behavior in this problem domain. The system is deployed in a P2P set-top box (STB) simulated environment. Initial test results validate the effectiveness of VODCA to evaluate and enhance the CAS algorithm for increasing replication efficiency.
