Off-Street Vehicular Fog for Catering Applications in 5G/B5G: A Trust-Based Task Mapping Solution and Open Research Issues

One of the key enablers in serving the applications requiring stringent latency in 5G networks is fog computing as it is situated closer to the end users. With the technological advancement of vehicles' on-board units, their computing capabilities are becoming robust, and considering the underutilization of the off-street vehicles, we envision that the off-street vehicles can be an enormously useful computational source for the fog computing. Additionally, clustering the vehicles would be advantageous in order to improve the service availability. As the vehicles become highly connected, trust is needed especially in distributed environments. However, vehicles are made from different manufacturers, and have different platforms, security mechanisms, and varying parking duration. These lead to the unpredictable behavior of the vehicles where quantifying trust value of vehicles would be difficult. A trust-based solution is necessary for task mapping as a task has a set of properties including expected time to complete, and trust requirements that need to be met. However, the existing metrics used for trust evaluation in the vehicular fog computing such as velocity and direction are not applicable in the off-street vehicle fog environments. In this paper, we propose a framework for quantifying the trust value of off-street vehicle fog computing facilities in 5G networks and forming logical clusters of vehicles based on the trust values. This allows tasks to be shared with multiple vehicles in the same cluster that meets the tasks' trust requirements. Further, we propose a novel task mapping algorithm to increase the vehicle resource utilization and meet the desired trust requirements while maintaining imposed latency requirements of 5G applications. Results obtained using iFogSim simulator demonstrate that the proposed solution increases vehicle resource utilization and reduces task drop noticeably. This paper presents open research issues pertaining to the study to lead the way for future research directions.