SMETO: Stable Matching for Energy-Minimized Task Offloading in Cloud-Fog Networks

In order to minimize the total energy consumption of a cloud-fog network, one of the most essential challenges is the assignment of subtasks from the task node (TN) to suitable fog nodes (FNs). In this paper, we apply a many-to-one matching to deal with this problem. Specifically, we first introduce two concepts, Service Efficiency (SE) and Energy Efficiency (EE), as the indexes of the preference list (PL) of TNs and FNs, respectively. Then a stable matching algorithm for energy-minimized task offloading (SMETO) is proposed, which is comprised of two key components: (i) Deferred Acceptance Algorithm Based On Energy Efficiency (EEDA) and (ii) Energy-Minimized Task Allocation (EMTA). Algorithm (i) is an iterative procedure that matches TNs and helpers based on PLs. Algorithm (ii) minimizes the energy consumption in the network by allocating the subtasks to helpers according to the result of matching. Finally, numerical simulations indicate the stability and energy-minimization of our proposed SMETO.