Design and Performance Analysis of Position-based Routing Protocol Using Location Prediction Techniques For VANET

Abstract

Vehicular Ad Hoc Network (VANET) is an emerging paradigm and an upcoming reality in road transportation which is designed and developed to minimize the road accidents and fuel consumption by giving a prior alert on traffic condition and collision detection using Vehicle to Vehicle (V2V) and Vehicle to Roadside Unit (V2R) communication. VANET is said to be a subset of Mobile Ad Hoc Network (MANET) due to similar node characteristics such as self-configuring node and multihop routing for examples. Based on their similar characteristics, MANET routing protocols are used in VANET. However, at various perspectives, VANET and MANET differentiate each other such as power, speed and mobility. Hence, applicability of MANET topology based routing protocols should be re-evaluated thoroughly using IEEE 802.11p standard which is specifically designed and developed to be used with VANET communication. In addition, realistic mobility should be generated by considering urban and non-urban vehicular traffic. However, position-based routing protocols are more suitable than the topology based routing protocols as position-based routing protocols do not maintain topology of the network rather they use the vehicle position on the Earth’s surface. The vehicle positions are obtained from the satellite navigation devices such as Global Positioning System (GPS), Galileo, Compass and Glonass. The position-based routing protocols use vehicle position as a location id during routing. Generally, satellite obtained locations are prone to have error due to environmental effect and urban infrastructure. Due to this error, the real vehicle position would be 05-100m away from the GPS location, approximately which effect the application performance. Thus, the location accuracy of the vehicle is a prime concern in VANET which enhances the application performance of automatic parking, cooperative driving, routing etc. to give some examples. Hence, various location prediction techniques are proposed in the literature to minimize the location error. In order to address these issues, at the outset, this thesis aims to re-evaluate the applicability of topology based routing protocols in VANET, particularly, it evaluates the Ad Hoc On-demand Distance Vector (AODV) and Optimized Link State Routing (OLSR) protocols on IEEE 802.11p standard. In the evaluation, it uses two different road network scenarios, particularly a complex road network, which represents the city road network, having multiple crossroads and an intersection of two roads. The second contribution of this thesis aims to propose a location prediction algorithm which is designed by considering nonlinear vehicular movement, as speed of the vehicle in the city varies between 0 to 60 Km/h, due to traffic rules, driving skills and traffic density. Likewise, the movement of the vehicle with steady speed is highly impractical. Consequently, the relationship between time and speed to reach the destination is nonlinear and with reference to the previous work on location prediction in VANET, nonlinear movement of the vehicle was not considered. In addition, location error also effects the performance of position-based routing protocol in VANET. Thus, it proposes a location prediction algorithm for a nonlinear vehicular movement using Extended Kalman Filter (EKF). EKF is more appropriate contrasted with the Kalman Filter (KF), as it is designed to work with the nonlinear system. The efficacy of the prediction algorithm is evaluated on real and model based mobility traces for the city and highway scenarios. Further, prediction accuracy of the EKF is compared with the KF on Average Euclidean Distance Error (AEDE), Distance Error (DE), Root Mean Square Error (RMSE) and Velocity Error (VE) metrics. The third contribution of the thesis proposes to use KF and EKF based location prediction techniques into the position-based routing protocol which is named as prediction based position-based routing protocol. The inclusion of KF and EKF based prediction techniques into the routing protocol has the aim to minimize the location error to improve the routing performance. The performance of the prediction based position-based routing protocol is evaluated on Two-ray ground and Winner-II propagation models with different transmission ranges of 250m and 500m for the city and highway scenarios. In simulation, two different traffic environments such as heterogeneous and homogeneous traffic are used. The performance of the prediction based position-based routing protocol using KF and EKF prediction module is compared with Cross-layer, Weighted and Position-based Routing (CLWPR) protocol on the metrics of Packet Delivery Ratio (PDR), Average Delay (AD) and throughput. The fourth contribution of this thesis evaluates the location prediction based position-based routing protocol using KF on real time GPS traces for 500m transmission range. In addition, it evaluates KF by predicting advance location of a vehicle on a highway scenario and error removal capacity. Based on obtained experimental results, it is observed that topology based routing protocols are less effective for VANET, while EKF based prediction is more accurate than KF based prediction. The PDR and AD got improved with EKF and KF based prediction in position-based routing protocol.