A Hybrid Approach on Human Driver Behavior Prediction and Network Performance Evaluation in VANET

Authors

  • S. Cloudin, P Mohan Kumar, J. Arokia Renjit

Abstract

The Vehicular Ad Hoc Network (VANET), constructs a mobile network of fast moving vehicles as computing and communicating nodes. A network of wider range can be created by connecting the vehicles together in an adhoc manner. Ensuring the road safety by deploying accident prevention techniques is the prime goal of Intelligent Transport system (ITS), for which the network back bone is supported by VANET.  It is the fact that the mistake of human drivers is the major reason for road accidents. Human driving pattern impacts and influences the performance and behavior of the network, because the topology of the network is mainly dependent on the movement pattern of vehicles.  Driving behavior is being classified as normal driving, reckless driving, fatigue driving and drunken driving. This paper contributes towards the performance analysis on the mobility of the network which is influenced by the prediction of the human driver behavior, while driving. The development of a driver behavior detection system using Mega Trend Diffusion (MTD) and Support Vector Machine (SVM) and the network performance analysis of the Greedy Perimeter Stateless Routing (GPSR) protocol, using an integrated simulation environment is discussed in this paper

Keywords- VANET, driver behavior, mobility pattern, MTD, SVM, network performance,GPSR

Published

2020-12-24

Issue

Section

Articles