vFog: A Vehicle-Assisted Computing Framework for Delay-Sensitive Applications in Smart Cities

• Main Authors: Shah, Syed Sarmad, Ali, Muhammad, Malik, Asad Waqar, Khan, Muazzam A., Ravana, Sri Devi
• Format: Article
• Published: Institute of Electrical and Electronics Engineers 2019
• Subjects: QA75 Electronic computers. Computer science
• Online Access: http://eprints.um.edu.my/23486/; https://doi.org/10.1109/ACCESS.2019.2903302
• Institution: Universiti Malaya

The inception of the smart cities concept provides a compelling platform to support innovative applications. It provides distinctive view of cities, where mobile devices, pedestrians, and electronic gadgets can communicate with each other to build an effective urban environment to further improve the living standards. Similarly, the role of the Internet of Things (IoT) and vehicular computing has emerged due to smart cities. This is further complemented by edge and fog computing architectures. The emerging concept of vehicular fog computing has enabled the platform to support delay-sensitive applications and to reduce the workload on the backend networks. Vehicular fog computing is a paradigm that touches the boundaries of thinking vehicles as an infrastructures-as-a-service. The use of vehicles to provide computation on-the-move poses various challenges. The vehicles with onboard computing equipment can facilitate delay-sensitive applications. These vehicles can act as an edge device to reduce the load from a backbone network. However, due to continuous mobility, it is difficult to use traditional frameworks to distribute the computation task among vehicles. In this paper, we propose a framework termed vFog. The vFog is designed to provide computing facilities from nearby fog vehicles. The framework utilizes the onboard computing facility of vehicles without the support of roadside units (RSUs). Moreover, the proposed framework handles churn behavior and supports multi-hop communication to improve the task delivery ratio. The proposed framework allows researchers to benchmark their own task distribution algorithms over the dynamic vehicular networks. © 2013 IEEE.