Optimization Studies Based On Adaptive Feedback Thinking Caching System Shunt

Proxy cache is an important network technology, which can reduce duplicated data transmission and make responses to users faster. It provides transparent data caching service to users, and will not affect the existing network architecture. When proxy cache system is placed at the edge of user group, it can significantly reduce the network traffic. So using proxy cache in P2P networks; it can significantly reduce the bandwidth consumption and will make users have better QoS performance.Proxy cache has many excellent cache replacement algorithms, but they do not provide priority distinction between requests. There are bottlenecks in system service, so proxy cache is unable to serve the users who request for hot resources. The dominant features of P2P applications are large amount of data transmission and long connection holding time, which should be considered in the optimization. Therefore, the solutions and the optimization algorithms should be based on these P2P characteristics. They make P2P proxy cache served the users who request for hot resources, and ensure QoS of users who request for cold resources.Based on the project of P2P Proxy Cache optimization, this thesis focus on the solutions for system bottleneck. On the basis of P2P routing, this thesis studied the optimization scheme which used feedback control mechanism. There are three major aspects included in the scheme:1,In order to concentrate system services on the requests for hot resources, the scheme distribute user requests in routing level. The more processes which serve the request for hot resources, the better system performance can be enhanced. Therefore, P2P Proxy Cache should provide services to the requests for hot resources, while forwarding the requests for cold resources to data source.2,This thesis implemented an adaptive feedback control mechanism. The cached resources can significantly improve system throughput, but distribution structure of P2P Proxy Cache works without any reasonable use of cached information. Therefore, with the feedback information, the mechanism makes routing system can control the user requests effectively. 3,This thesis presented and implemented an optimization algorithm of comprehensive evaluation. Combining the optimization theory and the demand of time complexity, the algorithm adapts to the dynamic behavior of users. With reasonable determination of resources'attributes, this algorithm can distribute requests effectively.According to the above optimization, a feedback control system was designed and developed on the P2P Proxy Cache. In the experimental environment, the functions of feedback control system were tested, and also the optimization algorithm was simulated. Experiment results show that the system can adapt to the dynamic user behavior; simultaneously, the optimization algorithm of resource evaluation can enhance system capacity and improve system performance.