Date of Award

Spring 4-2024

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Information Systems (PhDIS)

First Advisor

Yong Wang

Second Advisor

David Zeng

Third Advisor

Jun Liu

Fourth Advisor

Bhaskar Rimal

Abstract

Internet of Things (IoT) is commonly utilized in domestic and industrial environments to automate various tasks. Due to this, an enormous amount of data is being generated and transmitted through IoT networks. These data may contain sensitive information depending on the context. Access control is one of the frontline security measures that any information system should adopt. The dynamic nature of the IoT requires access control policies should be able to adapt to their environments. However, it is very challenging to specify access control policies manually because of their dynamic nature. Current literature suggests the need for automating the process of policy generation. Machine Learning and Deep Learning techniques can enable the required automation. The main objective of this dissertation is to answer the following research questions: 1) How can we self-generate contextual access control policies for the Internet of Things during unforeseen situations? 2) What are the existing challenges while specifying dynamic policies for access control in IoT? 3) How realistic are the generated access control policies to be used in real-time situations? In this research, we proposed a mixed-method approach where we implemented and evaluated two baseline Tabular Generative Adversarial Network models. We evaluated the performance of the solution using two datasets, namely the CAV Policies and Amazon Access Logs datasets. We obtained different perspectives based on our experiments. The common findings that our results demonstrate are that the models were able to generate synthetic access control policies by training from the datasets, and the models were able to learn the background knowledge specified during training to generate policies without any constraint violation.

Share

COinS