Enhanced K-means Clustering of Tasks and Virtual Machines for Load Balancing in Fog Environment

In Fog computing tasks with different starting time, execution time, and length are assigned directly to Virtual machines (VM). The clustering and load-balancing (LB) mechanisms are vital in optimizing resource utilization and network performance, preventing bottlenecks, and enhancing overall efficiency. This paper proposes Enhanced K-means clustering of tasks and VM for LB at fog layer (EKCLB), in which the clustering of tasks based on priority, burst time, and clustering of virtual machines based on capacity will be done to reduce makespan and to enhance the utilization of resources, further tasks have to be assigned to virtual machines at the fog layer in smart cities scenario. Our proposed clustering with load balancing algorithm is implemented using the iFogsim2 simulator. Simulation findings indicate that the proposed clustering with load balancing algorithm outperforms the non-clustering approach in minimizing makespan and enhancing resource usage.