Key Technologies On A Novel In-network Storage Model Named Buffer Bank

Large-scale streaming media distribution services cause unprecedented pressures and challenges to existing Internet architecture. Researchers have proposed a variety of network storage solutions to maximize the efficiency of content distribution, such as CDN,P2 P cache, ICN and so on. All these solutions achieve their goals by deploying storage resources close to the end users to alleviate the pressure of the core network, as well as to improve the user experience(QoE).Draw on the advantages of P2 P service, which take the advantage of end-users’ resources, we proposed BufferBank. BufferBank aggregates end-users’ resources and provides scalable storage services to streaming media applications. However, the dynamic and arbitrary character of the end-user behavior brings challenges to the designation of BufferBank.Therefore, our research focus on the problem of data management and data reliability.Our contribution could be summarized as:(1)We showed a formal definition of distributed buffer management issue in BufferBank and pointed out that the core issue of buffer management would be the various buffer allocation strategies through a simple example, and analyzed the performance of these three buffer allocation mechanisms.(2)We designed an evaluation model based on BBD behaviour. Through this model,BBD can get a score which can reflect the reliability property.(3)We designed and achieved a BufferBank simulator to analysis and evaluate different buffer management strategies performance. The number of redundant data migration is used as an indicator to evaluate different buffer management strategy performance.Our work is mainly to provide a new way of thinking for the dynamic, universal scalable storage system in the internet. It can maxmize data reliability while improving resource utilization. The implemention of simulator BB-SIM provides a new idea of simulate network communication systems.