Abstract: Many past systems have explored with the problem is how to eliminate a novel end-to-end traffic redundancy by which improves network efficiency. Several of these systems operate at the application layer while the more recent systems operate on individual packets. In this paper, we present PACK (Predictive ACKs), a novel designed for cloud computing customers end-to-end traffic redundancy elimination (TRE) system. PACK’s main advantage is its capability of increase the offloading the cloud-server TRE effort very end client, thus minimizing the processing costs activate by the TRE algorithm. Like other techniques PACK does not require any external support to maintain client’s status. By combing cloud computing client the efficiency per physical hosted server is improved greatly, data and resources are hotfoot provisioned provided as standardized offerings to users with a flexible price. But it is important to provide the convenient pricing model for the users of cloud. Hence we design a new traffic redundancy and elimination scheme for reducing the cloud bandwidth and costs. PACK is based on a novel TRE technique, which allows the client to use newly received block to identify previously received block chains, which in turn can be used as reliable, with respective predictors to future transmitted blocks. We present a fully functional PACK implementation transparent to all TCP-based applications and network devices. Finally, we analyze PACK benefits for cloud users, using traffic traces from various sources.
Keywords: cloud computing, traffic redundancy elimination, quality of service, caching, network optimization.