A Load-Balanced Task Scheduling in Fog-Cloud Architecture: A Machine Learning Approach

The swift expansion of internet-of-things (IoT) devices and the rise in the pace of task requests sent from these IoT devices to the cloud data centres led to Congestion and delays in the service. To meet the challenges, fog computing emerged as a new computer paradigm that offers services near the request-generating devices and reduces delays, particularly for real-time and delay-sensitive queries. It is crucial to consider issues like balancing the load, lowering energy consumption, and scheduling requests that impact the fog-cloud ecosystem's performance to accomplish these aims. This work suggests a Machine Learning based Task scheduling algorithm with load balancing for the fog-integrated cloud. It first deals with the task offloading to decide the layer where the service should be placed in the fog-cloud ecosystem. Then, it allocates the best available node considering the load balance of the overall ecosystem. The simulation experiments show that the suggested algorithm better balances the load and decreases reaction time compared to the state-of-art algorithms. It is also energy efficient as it minimises the number of active devices and their run time.