Detection of injection attacks on in-vehicle network using data analytics
Date
Authors
Major Professor
Advisor
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Authors
Research Projects
Organizational Units
Journal Issue
Is Version Of
Versions
Series
Department
Abstract
We investigate the possibility of detection of injection attacks using data analytics techniques
in this thesis. The automotive industry is innovating the modern vehicles towards connectivity by
interfacing them with various external entities. These entities are exposing the automobile to cyber
attacks instead of ensuring its safety. Therefore it is important to consider the security aspect while
developing these interfaces. Firstly, we try understand the automobile network architecture and the
possible security threats associated with it. Next, we examine the various possible cyber-attacks
on automobiles described in the literature. We experiment and analyze the attack scenarios by
performing injection attacks on a vehicle. We collect the data during the injection attacks and
apply multiple data analysis techniques. These techniques build a model based on data during
normal operation. The observations from the data collected during injection attacks is fit into
these techniques. The data points that do not fit the model are termed as attack points. Finally
we examine and analyze the results and their accuracy in detecting injection attacks.