Vehicular Fog Computing: Challenges Applications and Future Directions

Introduction

Mobile Ad Hoc Networks (MANETs) (Chlamtac, 2003; Conti & Giordano, 2014) are a collection of wireless devices like mobile phones, laptops and iPads that can dynamically form a wireless network for communication in emergency situations, disaster recovery operations (Menon et al., 2016) battlefields etc. These networks are deployed without the support of any fixed infrastructure like access points and does not have any centralized control. It is an autonomous system of mobile device in which every device can join or leave the network at any time leading to a highly dynamic and unpredictable topology. Vehicular Ad Hoc Networks (Li & Wang, 2007; Dua, 2014; Chen et al., 2015) are specific type of MANETs in which the mobile nodes are moving vehicles. In VANETs every vehicle is equipped with an on-board unit and a group of sensors. The radio interfaces or on-board unit enables short range wireless ad hoc networks to be formed. VANETs often has multiple Road Side Units (RSUs) deployed as intermediary servers near the vehicles to process the data. VANETs offer both vehicle to vehicle and vehicle to roadside unit communication (Harsch et al., 2007). Every vehicle in the network plays the role of a sender, receiver and a router to broadcast data to the vehicular network and the roadside units which then uses the data to ensure safe and free flow of traffic. VANETs are used in the design and development of Intelligent Transportation Systems (ITS) that offers improved safety and better transportation. VANETs are currently being used for traffic monitoring, emergency services, safe driving, infotainment services, location detection services, automated toll payment etc.

VANETs have been depending primarily on cloud computing (Vouk, 2008; Rountree & Castrillo, 2014) services for communication, computing and storage facilities. Cloud computing offers computation and storage facilities using a central cloud server or a group of remote servers. Cloud computing provides users with scalable virtual networks, virtual servers for remote storage space and computing facilities. Data stored can be accessed from any place without the trouble of keeping large storage and computing devices in the vehicles. Users could share and distribute large amount of data between the vehicles. One of the major areas of concern in cloud computing is the delay in transfer of data and information from the vehicles to the remote cloud server and back to the vehicles after storage and processing. With tremendous rise in the number of connected vehicles and with their ever-increasing mobility, the demand for applications that support low latency, uninterrupted services is rising day by day. It has become quite difficult to meet the challenges of efficient communication and computation with the emergence of latest and advanced vehicular applications. Providing required Quality of Service (QoS) is another important challenge faced by Vehicular Cloud Computing (VCC) (Whaiduzzaman et al., 2014; Shojafar et al., 2016; Mekki et al., 2016). Various vehicular applications designed for latest high-speed vehicles will require powerful communication and computational support. Apart from cloud computing, the other solutions used to provide communication and computational support was Fourth Generation Cellular Networks (Hampel et al., 2003) and Road Side Units. However, these technologies suffered from many limitations. Cellular networks were controlled by their respective service providers, thus limiting the flexibility of communication in VANETs (Hampel et al., 2003). Also, there were many difficulties in deploying Road Side Units at a larger scale (Kuo et al., 2013).