Major
Computer Engineering
Anticipated Graduation Year
2021
Access Type
Open Access
Abstract
The main objective of this research is to identify security threats that stem from autonomous vehicles on the road. To collect this research, I operated a smart car that utilizes ultrasonic sensors, light sensors, and remote control for semi-autonomy. These differing controls help to understand the presence of vulnerabilities. In addition, I explored how to improve the security of autonomous vehicles by implementing learning-based algorithms from Matlab-Simulink.
The Simulink model is based on neural networks and deep learning. A neural network is a computing model that resembles a networked structure of neurons in the brain. It allows for systems to learn from data in order to recognize patterns, classify data, and forecast future events. Deep learning refers to a neural network of over 100 layers. It often involves complex identification applications such as face recognition, text translation, and voice recognition.
This research attempts to use unsupervised machine learning in order to recognize the risks that come from autonomous vehicles. Unsupervised learning learns from past data that is not labeled, classified, or categorized and uses it for future events. Unlike supervised or reinforcement learning, unsupervised learning does not need to be programmed or instructed on how to identify differences. The levels of vulnerability are identified by their causes and effects to the normal functioning of the vehicle in question.
Faculty Mentors & Instructors
Dr. Brooke Abegaz, Assistant Professor of Computer Engineering, Engineering Science
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.
Improving the Security of Autonomous Vehicles using Unsupervised Machine Learning
The main objective of this research is to identify security threats that stem from autonomous vehicles on the road. To collect this research, I operated a smart car that utilizes ultrasonic sensors, light sensors, and remote control for semi-autonomy. These differing controls help to understand the presence of vulnerabilities. In addition, I explored how to improve the security of autonomous vehicles by implementing learning-based algorithms from Matlab-Simulink.
The Simulink model is based on neural networks and deep learning. A neural network is a computing model that resembles a networked structure of neurons in the brain. It allows for systems to learn from data in order to recognize patterns, classify data, and forecast future events. Deep learning refers to a neural network of over 100 layers. It often involves complex identification applications such as face recognition, text translation, and voice recognition.
This research attempts to use unsupervised machine learning in order to recognize the risks that come from autonomous vehicles. Unsupervised learning learns from past data that is not labeled, classified, or categorized and uses it for future events. Unlike supervised or reinforcement learning, unsupervised learning does not need to be programmed or instructed on how to identify differences. The levels of vulnerability are identified by their causes and effects to the normal functioning of the vehicle in question.