Adaptive Transfer-Based Fog Load Balancing Using Transfer Learning for Fault-Tolerant Cloud Systems

Fog load balancing is dynamic in improving the receptiveness and efficiency of cloud based distributed systems by task offloading in the vicinity of data sources. Current methods are not prone to dynamically adjust in the case of a slow or loaded load balancing node, resulting in system inefficiencies and possible downtimes. In a bid to overcome this lacuna, this paper introduces a new method named Adaptive Transfer-Based Fog Load Balancing (ATFLB). The proposed research work implements ATFLB for finding able alternative systems in case of failure and restructuring tasks smoothly. Through the utilization of transfer learning, the method facilitates efficient data transfer between nodes as well as ensures the delivery of load balancing metadata that is essential and utilized for best task distribution. This method allows for improved system performance, fault tolerance and reliability in fog-cloud environments.