Optimizing the energy consumption in three-tier cloud–edge–fog federated systems with omnidirectional offloading

Most of the recent studies regarding a collaborative federation architecture that integrates multiple computing systems to support Internet of Things (IoT) networks only consider two-tier cloud–edge federation systems, which only support vertical and horizontal offloading. This paper proposes a generic architecture of cloud–edge–fog systems, which provides omnidirectional offloading between service nodes. To investigate the effectiveness of the proposed architecture, it is formulated as a workload offloading and capacity optimization problem, to minimize energy consumption. The essence of our approach lies in harmonizing the unified computing resources across the cloud, edge, and fog layers, thus optimizing the system's overall energy efficiency and responsiveness. This architectural innovation is grounded in the philosophy that a more flexible workload offloading mechanism can significantly enhance the system's adaptability to dynamic workload demands, leading to substantial energy savings and improved service latency. Modified simulated annealing (MSA) is implemented to develop the offloading algorithm, which leads to a reduction of energy consumption of 11%–27% when compared to other heuristic methods. Simulation results also show that three-tier cloud–edge–fog federated architecture with omnidirectional offloading can reduce the total energy consumption by 12%–21% when compared to other existing architectures in the literature. We also get better results for average response time when compared with existing architectures with the same number of input workloads.