Load Balancing in a Heterogeneous Cloud Environment with a New Cloudlet Scheduling Strategy

The idea of cloud computing has completely changed the digital world in the age of the internet. Cloud computing offers a variety of cloud services, including software (Software as a Service), platforms (Platform as a Service), and infrastructure (Infrastructure as a Service). Pay-as-you-go is the primary business model used by cloud service providers (CSP). Services are offered that are in demand. Therefore, it is the only obligation of the cloud service provider to guarantee a terrific, continuous, and seamless service to its clients. To ensure an increase in service quality, load balancing is highly solicited. This paper mainly focused on task scheduling in a heterogeneous cloud environment so that no VM gets overloaded or under loaded. Users of cloud services typically submit tasks to CSPs. We mainly focused on enhancing the tasks’ completion rates and turnaround times. Enhancing the overall completion time of all jobs unquestionably contributes significantly to raising system throughput. This is accomplished by assigning a specific task to a particular virtual machine so that all machine loads are nearly equal and all tasks have nearly equal priority. Finally, we compare our task scheduling algorithm by substituting the original task scheduler with the one we have proposed. We then compare the results with the load-balancing techniques currently in use. The results are quite encouraging and have significantly improved the load balancer’s efficiency.