Date of Award

2018

Document Type

Open Access Master's Thesis

Degree Name

Master of Science in Electrical Engineering (MS)

Administrative Home Department

Department of Electrical and Computer Engineering

Advisor 1

Aurenice M. Oliveira

Committee Member 1

Glen E. Archer

Committee Member 2

Zhaohui Wang

Abstract

In recent years, connected vehicle technologies have been developed by automotive companies, academia, and researchers as part of Intelligent Transportation Systems (ITS). This group of stakeholders continue to work on these technologies to make them as reliable and cost-effective as possible. This attention is because of the increasing connected vehicles safety-related, entertainment, and traffic management applications, which have the potential to decrease the number of road accidents, save fuel and time for millions of daily commuters worldwide.

Vehicular Ad-Hoc Network (VANET), which is a subgroup of Mobile Ad-Hoc Network (MANET), is being developed and implemented in vehicles as the critical structure for connected vehicles applications. VANET provides a promising concept to reduce the number of fatalities caused by road accidents, to improve traffic efficiency, and to provide infotainment. To support the increasing number of safety-related applications, VANETs are required to perform reliably. Since VANETs promise numerous safety applications requiring time-bound delivery of data packets, it is also necessary to replicate real-world scenarios in simulations as accurately as possible.

Taking into account the effect of realistic obstacles while simulating a variety of case scenarios increases the reliability of the tested routing protocol to appropriately perform in real-world situations. It also exposes routing protocols to possible vulnerabilities caused by obstacles. Nevertheless, it is not uncommon for researchers to omit real-world physical layer communication hurdles in simulation-based tests, including not considering the effect of obstacles on their routing protocol performance evaluation simulations. Consequently, the performance of these protocols is usually overestimated and do not support in real-world environment. Failure to account for obstacle effects overstate the network performance. In this thesis, a framework for measuring obstacle effects on routing protocols is defined. We also propose, a new routing protocol based on the traditional Greedy Perimeter Stateless Routing (GPSR) protocol called Intelligent Greedy Routing (IGR) protocol.

The proposed IGR protocol considers a parameter called $Receptivity$ to chose the next hop in a route. We implemented the new protocol using the Simulation of Urban Mobility (SUMO) and the Network Simulator (NS-3). An analysis of Packet Delivery Ratio (PDR), End-to-End Delay (E2ED) and Mean Hop count with the assumption that nodes (vehicles) are moving in various topologies is presented in this thesis. The study presented here gives a general idea of the effects of obstacles on the Greedy Perimeter Stateless Routing (GPSR) protocol considering multiple realistic scenarios such as Urban, Residential and Highway. In addition, we compare the performance of GPSR and the new IGR protocols with the presence of obstacles considering various topologies. The new proposed IGR protocol performs better compared to the traditional GPSR for all the investigated metrics.

Download

Share

COinS