Optimizing Cloud Bandwidth and Cost using Content Aware Chunk Prediction to Improve Efficiency of Retrieval

Abstract

Bandwidth, an important most demanding computing resource is mainly focused on backup service in cloud computing technology. In addition to backup solutions, communication and transferring of data also involves a considerable amount of bandwidth for efficient throughput. Management of wide area network bandwidth provision and cost is a major challenge in recent periods. Due to enormous content in cloud, cost effective bandwidth provisioning solution should be adopted by the cloud providers for better user satisfaction. Here we use a min chunk hashing technique based on the content for preserving all chunks in order to reduce the overall transfer time for individuals as well as to reduce the resource allocation load on the cloud servers. Thus, content based chunking along with the metadata considerably reduces the cost of bandwidth utilization for individuals. The bandwidth usage is predicted using cloudsim simulator through BwProvisioner class provided in the cloudbus package. A sample of large text files with redundant data is refined, chunked and hashed using Message Digest5 to measure the transfer time and cost performance metrics for experimental values in a dynamic work load environment.