A GWO-MFO-Based Resource Allocation in Vehicular Fog Computing With Latency Constraints and Energy Reduction

The increase in general-purpose vehicular applications has driven the demand for robust low-latency communication and processing capabilities in Internet of Things (IoT) networks, especially in devices with limited resources. For this purpose, the establishment of Vehicle Fog Computing (VFC) is recommended. The processing, storage and computing capacity of intelligent vehicles in parking lots can be used for important applications as fog nodes. In this paper, the Workflow Scheduling and Resource Allocation (WSRA) as a multi-objective optimization problem is formulated. The GWO-MFO algorithm, incorporating both Gray Wolf Optimization (GWO) and Moth-flame Optimization (MFO), is intended to optimize the reduction of latency and energy consumption, while still meeting deadline and reliability requirements. GWO-MFO searches the problem search space more powerfully and obtains the best solutions. Minimization of objective function in GWO-MFO, using Processor Merging (PrM) and Dynamic Voltage Frequency Scaling (DVFS) techniques, reduces static and dynamic energy consumption, respectively. The effectiveness of the proposed algorithm was checked on three workflows of 6, 14, and 15 tasks based on the evaluation metric of energy consumption and latency. According to the experiments, the proposed method achieves better performance (latency and energy consumption) than the compared methods.