A dynamic load balancing mechanism for fog computing environment

Fog computing has emerged as an extension of cloud computing that provides cloud-like services at the edge of the to internet of things applications. However, the limited storage and processing capability of fog nodes along with inefficient resource scheduling creates a performance bottleneck that can result in high latency and network bandwidth. Therefore, dynamic load balancing is necessary to achieve the true benefits of fog computing. This paper proposes dynamic load balancing mechanism (DLBM) to schedule the number of service requests on fog nodes effectively. Furthermore, three algorithms to dynamically balance the load of fog nodes named are appropriate node selection, effective task distribution and global task execution and resource allocation. The performance of our proposed mechanism is compared with cloud only technique, fog-cloud-placement algorithm, and self-similarity-based load balancing technique. Comparative performance analysis validates the efficiency of the proposed approach. DLBM demonstrates considerable reduction in latency and network bandwidth utilisation.