SMRETO: Stable Matching for Reliable and Efficient Task Offloading in Fog-Enabled IoT Networks

Abstract

Fog computing is a key technology that supports timely and efficient computation of different tasks in IoT networks. By using the nearby fog nodes for quick task computation, application related decisions by IoT devices can be taken within the delay requirements. Resource allocation in terms of task placement on the free computing resources of the fog nodes is a major challenge in IoT networks. In this paper, we consider task offloading from IoT devices to the logically partitioned fog computing resources known as Virtual Resource Units (VRUs) to reduce the number of task outages and energy consumption of the IoT and fog nodes. We propose a two phased task offloading algorithm to minimize the number of task outages. In the first phase, we utilize the task deadline to compute the minimum number of resources required for a task from the fog nodes. To meet the heterogeneous task computing requirements, we introduce the concept of variable sized VRUs in the fog nodes. Moreover, we propose a modified Deferred Acceptance Algorithm (DAA) for stable matching between IoT tasks and variable sized VRUs. In the second phase of the algorithm, the unmatched fog node resources are distributed among the previously matched IoT tasks. Simulation results show that the proposed algorithm outperforms available techniques in the literature in terms of task outages and energy efficiency.