Research Of Management Protocol In Mesh-Based P2P Real-Time Streaming Media

At present, P2P streaming media technology has become the most widely used streaming media applications technology on the Internet.Therefore,the business of streaming media based on P2P has developed quickly, the business types and users in P2P streaming media system are becoming more, user management,business management and business retrieval are becoming more complicated.Furthermore, it is easy to appear a lot of latent selfish nodes which only share resources without contributing resources to others.These uncertain factors make it easy to appear users' behavior and media resources chaotic situations and nonstandard of business management,which may impact business decision makers to bring economic losses or even more seriously, making system performance worse and resulting in system crash.In this paper,firstly,we describe the research background and significance of the subject, the development status of P2P streaming media at home and abroad, and theoretical knowledge related.Then, analyse the public features of users behavior in current P2P streaming media systems,that is most users send their requests according to their fun,we propose a clustering algorithm based on users'interest that strengthens the users'management.Secondly,in order to reduce the media resources chaos caused by numerous business,we design intelligent program list based on standards of business decision makers.While personalizing programs navigation based on users'interests which makes business retrieval easier.Thirdly,we compare and analyze incentive mechanisms commonly used for P2P streaming media,propose incentive approach of trust model based on reputation to encourage selfish nodes to collaborate and contribute resources for the system.Based on the above,we conduct the overall design of the management protocol of mesh-based P2P streaming media system,the main content of the management protocol includes the definition of data structures and interfaces,the realization of clustering algorithm based on users'interests and intelligent program list,the implement of personalized programs navigation based on users'interests,and incentive approaches based on trust model of reputation.location of the user node's physical attributes.design and implementation of the users'contribution ability algorithm,the design and implementation of TOA,media pricing mechanism, and users'stability and level settings. Finally, we develope experiment platform for testing management protocol,we analyze and summarize the experimental results.