Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy (PhD)

Degree Program

Electrical and Computer Engineering

Year Degree Awarded

2017

Month Degree Awarded

May

First Advisor

Hossein Pishro-Nik

Second Advisor

Dennis Goeckel

Third Advisor

Marco Duarte

Fourth Advisor

Daiheng Ni

Subject Categories

Electrical and Computer Engineering | Systems and Communications

Abstract

This dissertation studies the ability to individualize vehicular ad hoc networks (VANETs) in order to improve safety. Adapting a VANET to both its individual drivers' characteristics and traffic conditions enables it to transmit in a smart manner to other vehicles. This improvement is now possible due to the progress that is being made in VANETs.

To accomplish this adaptation, our approach is to use VANET data to learn drivers' characteristics. This information along with the traffic data, can be used to customize the VANETs to individual drivers. In this dissertation, we show that this process benefits all the drivers by reducing the collision probability of the network of vehicles. Our Monte Carlo simulation results show that this approach achieves more than 25% reduction in traffic collision probability compared to the case with optimized equal vehicular communication access for each vehicle. Therefore, it has a considerable advantage over other systems.

First, we propose a method to estimate the distribution of a driver's characteristics by employing the VANET data. This is essential for our intended application in accident warning systems and vehicular communications.

Second, this estimated distribution and the traffic information are used to adapt the transmission rates of vehicles to each driver's safety level in order to reduce the number of collisions in the network. We derive the packet success probability for a chain of vehicles by taking multi-user interference, path loss, and fading into account. Then, by considering the delay constraints and types of potential collisions, we approximate the required channel access probabilities and illustrate the collision probability.

Third, since the packet success probability and thus communication interference affect the collision probability noticeably, we examine various interference models and their effect on the collision probability with more scrutiny. In our analysis, two signal propagation models with and without carrier sensing are considered for the dissemination of periodic safety messages, and it is illustrated how employing more accurate interference models results in a higher level of safety (lower collision probability)for the network.

Finally, there is an unclear relation between the intensity of an ad hoc network (the number of vehicles in a certain area) and the performance of the system. Hence, we study a reverse approach in which the geometry (intensity) of the unmanned aerial vehicles varies and certain requirements such as safety and coverage need to be satisfied. The numerical results show that safety and interference limits the coverage of the network and there is only a relatively small range of intensities which satisfy all three.

Share

COinS