A Many-to-One Matching based Task Offloading (MATO) Scheme for Fog computing-enabled IoT Systems

Fog computing networks have been widely integrated in IoT-based systems to improve the quality of services (QoS) such as low response service delay through efficient offloading algorithms. However, designing an efficient offloading solution is still facing many challenges including the complicated heterogeneity of fog computing devices and complex computation tasks. In addition, the need for a scalable and distributed algorithm with low computational complexity can be unachievable by global optimization approaches with centralized information management in the dense fog networks. In these regards, this paper proposes a distributed computation offloading framework (MATO) for offloading the splittable tasks using matching theory. Through the extensive simulation analysis, the proposed approaches show potential advantages in reducing the average delay significantly in the systems compared to some related works.