Efficient Task Offloading in NOMA-Assisted Vehicular Fog Computing for Customer Applications

Future Fog computing networks are the basis of many smart customer applications in the areas of transportation and healthcare. Timely execution of application related tasks is a key challenge in such fog computing networks so that application reliability and customer satisfaction can be maximized. In this paper, we focus on vehicular fog computing scenario and utilize Non-Orthogonal Multiple Access (NOMA) to improve the spectral efficiency of task transmission from vehicles to fog nodes. To tackle the NOMA pairing problem, we propose a Hungarian-based algorithm to minimize task computational delay. We consider multiple factors such as task priority, vehicle to fog node transmission rate, and current load at the fog nodes to optimize the pairing of vehicles for task transmission. For multi-criteria decision making, we utilize Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) to evaluate the ranking scores that are used in the Hungarian algorithm. Realistic simulation results for the proposed technique shows improved task computational delay for high priority tasks as compared to other techniques in the literature.