Prediction System for Reducing the Cloud Bandwidth and Cost

In this paper, we present AACK (Anticipating ACKs), a novel end-to-end traffic redundancy elimination (TRE) system, designed for cloud computing customers. Cloud-based traffic redundancy elimination needs to apply a judicious use of cloud resources so that the bandwidth cost reducing is combined with the additional bandwidth cost of traffic redundancy elimination computation and storage would be reduced. AACK’s main advantage is its capability of offloading the cloud- server traffic redundancy elimination effort to end-clients, thus minimizing the processing costs induced by the traffic redundancy elimination algorithm. Unlike previous solutions, AACK does not require the server to continuously maintain clients’ status. This makes AACK very suitable for pervasive computation environments that combine client mobility and server migration to maintain cloud elasticity. AACK is based on a novel traffic redundancy elimination technique, which allows the client to use newly received packets to identify previously received packet chains, which in turn can be used as reliable predictors to future transmitted packets. We present a fully functional ACK implementation, transparent to all TCP-based applications and net-work devices. Finally, we analyze AACK benefits for cloud users, using traffic traces from various sources.