Effective Resource Provisioning and Caching with Dynamic Changes by using Self Learning of Resources

Cloud based Content distribution networks are most famous among the small scale content distribution network organization. Because it enables an organization to make use of required storage, bandwidth and CPU capacity from the cloud services instead of having to buy and maintain those. In the previous work resource provisioning and caching contents in the dynamic environment is proposed to allocate the resources for storing the contents. However, the previous works doesn’t concentrate about the varying content file sizes as well as providing worst resource allocation at the time of more no of demands received from the users. In this work distributed learning algorithm is proposed to effectively handle the dynamic varying of user demands, and time varying user demands. Reinforcement learning is mainly used for adaptive management of Virtual machine capacity. Markov decision process is used to model the reinforcement learning process. The experimental results prove that the proposed work provides better result than the existing approaches.