Show simple item record

dc.contributor.advisorSalinas Monroy, Sergio A.
dc.contributor.authorRott, Michael
dc.descriptionThesis (M.S.)-- Wichita State University, College of Engineering, Dept. of Electrical Engineering and Computer Science
dc.description.abstractThe introduction of the Internet of Things (IoT) has helped improve efficiency and productivity through increased connectivity between devices. At the same time, this increased connectivity has opened new avenues for an attack against previously air-gapped systems. The malicious alteration of products or processes via cyber-attack is a real and present danger for the manufacturing industry that can lead to catastrophic conclusions, not just to the reputation and bottom line of a corporation, but also to the safety of employees and consumers. This research proposes machine learning software would allow manufacturers to proactively mitigate the possibility of an attack that has become all too real in a world full of remote attacks. The focus of this work is on cyber-attacks against additive manufacturing, also called 3D-printing. It proposes a novel intrusion detection approach that can detect defect injection (DI) attacks based on analyzing the 3D-printer’s power consumption. Existing intrusion detection techniques are designed for Information Technology (IT) systems and ignore attacks that compromise the electronic and physical components of 3D-printers. In contrast, this approach monitors the 3D-printers’ power consumption to detect malicious intruders using a deep-learning multi-layer neural network that predicts the average peak current consumed during each printer movement. If the observed measurement differs from the predicted by more than a specified threshold, then it is likely that an intruder is maliciously manipulating the 3D-printer. This system was used to test several attacks against a 3D printer testbed. The results show an accuracy of attack detection greater than 98% in detection of insertion, deletion, and void attacks. This meets or exceeds several previously researched methods.
dc.format.extentix, 35 pages
dc.publisherWichita State University
dc.rightsCopyright 2020 by Michael Rott All Rights Reserved
dc.subject.lcshElectronic dissertations
dc.titleIntrusion detection for additive manufacturing: A side-channel analysis approach

Files in this item


This item appears in the following Collection(s)

  • Master's Theses
    This collection includes Master's theses completed at the Wichita State University Graduate School (Fall 2005 --)

Show simple item record