A task offloading strategy considering forwarding errors based on cloud–fog collaboration

Fog computing, specialized in handling latency-sensitive and resource-hungry tasks, emerges as a pivotal paradigm for Internet of Things (IoT). Task offloading encompasses judgment and forwarding. However, current research predominantly concentrates on the former, neglecting forwarding and the potential occurrence of errors in this process. Additionally, existing models commonly employ continuous-time queuing models. To overcome these limitations, we propose a task offloading strategy considering forwarding errors based on cloud–fog collaboration. The strategy aims to offload tasks according to a predefined offloading ratio. We incorporate forwarding error ratios for all tasks, prioritize access for latency-sensitive tasks, and devise a discrete-time queueing model. Through numerical experiments, we analyze performance trends with offloading ratio. Additionally, a system profit function is employed to ascertain the optimal offloading ratio balancing the key metrics. Our findings underscore the significant advantages of cloud–fog collaboration over traditional pure cloud computing, notably increasing throughput rate while decreasing the blocking rate.