英文摘要 | It was reported by China Internet Network Information Center (CNNIC) that the number of network user in China had reached 591 million and the number of online video users in China had increased to 389 million by the end of June 2013. The rapid growth of online video users brings both opportunities and challenges to streaming media service system operators. Since streaming media services have the characteristics of high bandwidth consumption, strict real-time transmission requirements, long session duration, et al, it’ s a big challenge to provide high quality of streaming media service for users. The cost of streaming media distribution of traditional CDN (Content Delivery Network) model increases linearly with the number of user, and the system performance bottlenecks still exist. P2P streaming media distribution technology in the application of streaming media services brought new changes. Nowadays, the streaming media distribution model which combines CDN and P2P has become one of the most economical and reliable model to provide services. CDN can guarantee the quality of streaming media service for users, meanwhile, P2P users can utilize their idle resources, such as storage, CPU, bandwidth, et al, to provide streaming media services for other users, which can reduce the load of CDN edge proxy server. How to better combine P2P with CDN, bring their advantages into full play , and build a safe, reliable, efficient and economical streaming media content delivery network is a challenging research direction, which has become a research hotspot. We study the key problems of the current P2P-based streaming media content distribution in depth, including how to maximize the utilization of user node’ s idle resources to serve the system in a heterogeneous network environment, and how to minimize users’ waiting time when they perform frequent VCR operation. By proposing an high efficient and scalable P2P streaming media content delivery architecture, called ORUVoD (Optimal Resource Utilizing Video on Demand), and based on this architecture, proposing an optimized pre-fetching scheme with guided seeks and cache algorithms based on time series prediction, we make full use of user node’ s idle resources in P2P-based streaming media content delivery network and effectively reduced users’ wait... |
修改评论