Multi-Objective Fuzzy Optimization for Energy, Carbon, and SLA Efficiency in Cloud Computing

Abstract

The fast development of cloud computing has contributed to the great proliferation of datacenters across the planet, which means the gluttonous use of electricity and releasing more carbon dioxide. These do not only increase the cost of operations but also promote degradation of the environment in the global environment. Ensuring Quality of Service (QoS) and ensuring energy efficiency in a cloud service environment is a significant challenge to cloud service providers. This is a multi-objective optimization problem whose objectives are conflicting in that minimization of energy consumption, minimization of carbon footprint, and performance of the SLA will not be achieved simultaneously. This study suggests a multi-objective fuzzy optimization model, which can be used to balance the energy use, carbon emission, and SLA efficiency in dynamic cloud system environments in an intelligent manner. The framework incorporates a fuzzy inference system (FIS) to control the uncertainty and vagueness of workload changes, server usage, and the environment conditions. The system calculates the trade-offs of the energy, carbon and the SLA parameters in real time based on the fuzzy rules and membership functions. A hybrid metaheuristic optimizer with the ability to execute a global search and a local search is then applied on the fuzzy result to determine the best virtual machine placement and task scheduling techniques. The proposed model is rather dynamic in regards to the different workloads and thermal conditions, unlike the more traditional methods used, which are based on a predetermined set of deterministic thresholds, which guarantee sustainable resource distribution. The framework will help to reduce energy consumption and carbon emissions and decrease the violations of SLA considerably by integrating smart decision-making and multi-objective balancing. The study helps in the development of green and sustainable cloud computing, as it provides an adaptable, versatile, and green optimization model that can be used in present-day datacenter operations.