EDGE COMPUTING APPLICATIONS ON VEHICULAR NETWORKS

Abstract

This paper investigates the various aspects of VANET (Vehicular Adhoc Networks) and usage of Edge computing to implement the same. In this paper we shall be looking into the basic challenges in VANET, the routing protocols in VANET, the types of wireless deployments and infrastructure typically used in VANET, the security challenges and the areas where Edge computing is used for VANET deployments. We shall also do a brief survey of 802.11p the wireless standard developed for VANET. VANET or vehicular adhoc network is a technology that enables communication and information sharing between moving vehicles. They form the backbone of Intelligent Transport Systems (ITS). The vehicles either communicate using an adhoc network set up among themselves or by communicating through an infrastructure. Such Infrastructure can be provided by installations on roadside or by some designated mobile nodes that are capable of decent computation capability. The information shared in these cases can be varied ranging from location of corresponding devices to other sensor data like picture of road obstructions, average traffic speed, locations of nearby traffic etc. The data thus collected can be fed to a backend server for analysis and information retrieval. Such information forms the basis of vehicle monitoring, traffic prediction and even self-driven cars. It has been observed that in most cases the data analysis is generally done at the backend as for analysis, data from multiple vehicles are needed and hence from an architecture point of view it is more logical to have these analyzed on some specific nodes in the network. In general, the data thus gathered needs to be analyzed and instruction/communication provided back to the VANET in real time. These make edge computing an ideal choice for such systems. Given the fact that the system needs to provide a decision based on the captured data, machine learning is an ideal choice in these cases. Such analysis should also take into the account the type of traffic and QOS parameters related to it. All the above put together have given rise to a new domain called Vehicular Edge computing.

Keywords: VANET, Edge Computing, Routing protocols, 802.11p, Network Security, Network Latency, Machine Learning, Cloud Computing, QOS, Vehicular Edge computing, Intelligent Transport Systems (ITS)