A Survey on Neural Network Based Minimization of Data Center in Power Consumption

Cloud computing is the current technology used for sharing and accessing resources via internet. Reducing power consumption is an essential requirement for cloud resource providers to decrease operating costs. We employ the predictor to predict future load demand based on historical demand. According to the Prediction, the algorithm turns off unused servers and restarts them to minimize the number of running servers. Power consumption can be regulated by using proper load balancing technique. Load balancing is done so to distributing the load fairly amidst the servers and also a scheduling technique is followed to selectively hibernate the servers to optimize the energy consumption. The load balancing is based on load prediction and server selection policy. Neural Network is used for load prediction, which predicts the future load based on past historical data. The servers can be monitored and given ranking based on their reliability record and this information is used as a criterion while performing load balancing.