Performance Optimization and Energy Minimization of Cloud Data Center Using Optimal Switching and Load Distribution Model

Cloud computing is an effective computing methodology used in all stages of business. Most of the Cloud Data Centers (CDC) operates on the basis of peak load and huge scales. Hence, it necessitates saving the energy in CDC. This study introduces an energy-efficient strategy based on the fat tree. Here, Taylor-based Manta-Ray Foraging Optimization (Taylor-MRFO) is developed by combining the Taylor series with Manta Ray Foraging Optimization (MRFO) to distribute the load in a CDC. In load distribution, the cloud data switching to the preferred mode is done by the Actor critic neural network (ACNN). Furthermore, the developed Taylor-MRFO+ACNN provided a better outcome than the conventional approaches with the least energy consumption of 0.4930, least load of 0.3631, and least fitness of 0.4343. For setup-1, when the population size is 15, the load value obtained by the proposed method is 23.43%, 10.19%, 7.18%, 5.31%, 4.43%, and 2.58% higher when compared to the existing approaches namely, Artificial Bee colony(ABC), Efficient Load Optimization and Resource Minimization (ELORM), Adaptive Parameter- Ant Colony Optimization (AP-ACO), Multi-Objective Memetic Algorithm-Adaptive Plant Intelligent Behavior Optimization (MOMA-APIBO), Cooling Control Algorithm (CCA), and Minimum Total Power (MinPR).