|
" Trajectory Privacy in Autonomous Vehicles Communications and Applications "
Bani Hani, Abdelnasser Ghaleb Mohammad
Fu, Huirong
| Document Type
|
:
|
Latin Dissertation
|
| Language of Document
|
:
|
English
|
| Record Number
|
:
|
1054211
|
| Doc. No
|
:
|
TL53328
|
| Main Entry
|
:
|
Bani Hani, Abdelnasser Ghaleb Mohammad
|
| Title & Author
|
:
|
Trajectory Privacy in Autonomous Vehicles Communications and Applications\ Bani Hani, Abdelnasser Ghaleb MohammadFu, Huirong
|
| College
|
:
|
Oakland University
|
| Date
|
:
|
2020
|
| Degree
|
:
|
Ph.D.
|
| student score
|
:
|
2020
|
| Note
|
:
|
110 p.
|
| Abstract
|
:
|
Current wireless networks are not designed to satisfy privacy concerns. In contrast, Dedicated Short Range Communications (DSRC) in Vehicular Ad Hoc Networks (VANETs) are designed and intended to achieve privacy. Consequently, any special identifications, such as mac addresses, IP addresses, and certificates, are designed to be changed from time to time to meet privacy requirements. Privacy requirements, which are neglected in traditional wireless networks, are taken into accounts in DSRC design. Different properties have been proposed and other properties are expected to appear while VANETs research advances in the future. Location Based Services (LBSs) become more ubiquitous due to modern and advanced technologies, such as Global Positioning System (GPS) and Fourth Generation Data Technology (4G) of cellular networks. Moreover, LBSs are expected to gain more attention with new technologies such as Fifth Generation of Data Technology (5G), Internet of Things (IoT), VANETs, and smart grids. Privacy as a citizen’s right has confounded consumers and legislators, more and more as technology seems to chip away of personal anonymity. Nowadays, in the dawning of the age of Autonomous Vehicles (AVs), engineers and businesspeople perhaps soon need to grapple with privacy as a quantifiable, marketable feature of automotive products. It is possible that one day automotive manufacturers may have to compete not only on price and gas mileage, but also on the degree to which vehicles protect users’ information. Location privacy measurement is complex. It is necessary to consider the protocols involved. If the LBSs use DSRC, then it is possible to implement Pseudo Identifiers (Pseudo-IDs) for each car. Different approaches have been proposed to measure location privacy, such as snapshot (point in time) privacy or continuous location privacy [1, 2]. Sometimes, distance between anonymized vehicles is a factor [2]. However, AVs could implement diversion tactics, redirecting in mix-zones, tactics which humans might find too difficult to execute. In order to measure the difference, a clear definition and metric for trajectory privacy are needed. This research provides a systematic study of autonomous vehicle location privacy, with a focus on trajectory privacy. Past research has evaluated protections against "tracking" vehicles, such as deanonymizing vehicles in real time. This research evaluates protections against "tracing" vehicles, including deanonymizing vehicles, based on historical records of entire paths (trajectories). A recent work [3] suggests that there may exist previously unknown trajectory privacy methods peculiar to autonomous cars, and there may even be a need to update privacy metrics to better evaluate new privacy methods. Briefly, this work makes the following contributions to the ongoing field of location privacy: 1. Propose and define a new privacy model, which called Trajectory privacy by utilizing AVs Characteristics. 2. Design, Implement and Evaluate a new trajectory privacy attack on autonomous driving, called T-PAAD. 3. Propose a new performance metric to precisely evaluate trajectory privacy in order to illustrate the existing metrics insufficiency. 4. Design, Implement and Evaluate a countermeasure to mitigate the impacts of T-PAAD attack.
|
| Descriptor
|
:
|
Computer science
|
|
|
:
|
Information science
|
| Added Entry
|
:
|
Fu, Huirong
|
| Added Entry
|
:
|
Oakland University
|
| |