Date of Award

5-2019

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Computer Engineering and Sciences

First Advisor

Siddhartha Bhattacharyya

Second Advisor

William H. Allen III

Third Advisor

Veton Z. Kepuska

Fourth Advisor

Philip J. Bernhard

Abstract

The Internet of Things (IoT) has provided a flexible platform for a large number of heterogeneous devices to dynamically join and leave the network. This enhances the availability of a diverse range of services provided by a network. However, this dynamic expansion of the network with mobile IoT devices introduces a major challenge to security especially related to management of trust across the IoT platforms. Furthermore, IoT is open and distributed in nature, which allows the integration and registration of diverse entities. Thus arises the necessity for a mechanism that can ensure the selection of secure and trusted devices as these devices try to join the network. In this work, an effective trust modeling mechanism has been investigated to support the development of trust based on a three-tier trust architecture. Initial trust establishment is done based on the notion of social trust semantics and personal attributes. This is represented as the first-tier of the trust architecture. In the second tier, the recommendation from another known device is taken into account to support the trust-relationship establishment process. The third tier shared (common) personal attributes could be useful to evaluate shared list of attributes among the participating devices. The first two tiers of the proposed three layered trust architecture has been implemented in a knowledge representation environment using ontology. Satisfaction of the trust properties in the proposed ontology is validated by SWRL rules. To enhance the functionality of trust methodology the ontology was mapped into the implementation environment for automated reasoning. To model and analyze the dynamic interaction between the IoT devices it considers various social trust properties such as honesty, dishonesty, legitimate request and trustworthiness to categorize device behavior. Finally, the performance of the trust assessment algorithm implemented in the proposed reasoner was validated by implementing several scenarios.

Download

Share

COinS