Optimize The Deployment Of P2p Content Distribution Network-based Data

Nowadays, with the development of the Internet, streaming video content is increasing, the Internet has become the main bearer of information. Compared with the traditional text, pictures and other traditional WEB applications, streaming media with high data volume, high-bandwidth, high traffic and the characteristics of the high quality of service requirements, and current Internet " best effort " to determine the characteristics of the existing network infrastructure under the existence of large-scale streaming media applications are difficult.In this case, CDN technology and P2P technology came into being, they are in different ways and different strengths to solve the problem of content distribution, but they are constrained by its inherent computational model, whether it is P2P technology or CDN technology are there some fundamental shortcomings. As the current two main stream content delivery technology, P2P and CDN technology and technical strengths, their differences in the calculation model has a complementary, if combined with P2P and CDN's scalability and manageability, reliability can build a managed, carrier-class content to the application of streaming media content distribution platform. This article will explore the use of CDN and P2P technology, complementary, and play to their strengths, on the P2P and CDN technology integration technology and built a new type of streaming media content distribution model in the deployment of research data, this thesis will study the following four aspects, test the proposed algorithm and verify the relevant.(1) load balancing between nodesNode through the physical and logical neighbors by neighbors to design the load balancing algorithm, and simulation verification.(2) the optimal allocation of storage spaceThis article from the storage space of the cache mechanism, a simple and practical buffer algorithm, and test validation.(3) optimal allocation of storage space for the active deploymentOptimal allocation model known number of copies, how to P2P distributed environment in a dynamic way to achieve the deployment, the node will be the optimal distribution of connectivity, document popularity, and many other effects, some model assumptions need to be in the implementation soft constraints; the other in a distributed environment there is no centralized node to obtain accurate information about the overall cost of expensive, heuristic rules will be superior to the traditional mechanisms of control mechanisms.(4) CDN active deployment of dynamic dataIn this paper, a dynamic deployment of data, that is the data dynamic scheduling, data partitioning algorithm is proposed and verified.