An Optimised Energy Efficient Task Scheduling Algorithm based on Deep Learning Technique for Energy Consumption

The information technology (IT) and mobile computing industries are now in the development stages of cloud computing (CC). Instead of being purchased, resources such as software, CPUs, memory, I/O hardware, and others are used and charged as needed. The massive expansion of CC necessitates enormous energy consumption, or data centers house a diverse spectrum of computers. Consequently, cloud service providers are exploring low-cost strategies for reducing energy use and carbon emissions. Therefore, work planning has garnered great attention and critical consideration about effective resources and bad energy consumption. This paper proposes a machine learning technique called short-term or Long-Term Memory (LSTM) for efficient power task scheduling to address growing carbon or energy emissions. The recommended strategy for scheduling considers the finish time or exclusive usage of a resource task, as well as the standardizing process. The Novel Black Window is used to reduce weight and improve the performance of LTSM. The simulated analysis is used to evaluate the efficiency of the LSTM-NBW in aspects of makes pan, power consumption, task completion time, and resource utilization. The findings show that the suggested model only obtained 400KWh more for the 80kB user job than the original LSTM model.