Committee: 

Dr. Beyah, Advisor        

Dr. Sarioglu, Chair

Dr. Qi

Dr. Zonouz

Dr. Gupta

Abstract: The objective of the proposed research is to secure Additive Manufacturing (AM) systems from the perspective of side-channels. On the one hand, we propose a practical Intrusion Detection System (IDS) leveraging side-channel signals in AM systems. On the other hand, we explore technical methods to fight against side-channel attacks to steal Intellectual Property (IP) in AM systems. For the IDS, we propose a novel framework called NSYNC to compare two side-channel signals in real-time. As of now, NSYNC is the only IDS that can efficiently compare two side-channel signals with time noise, which is the random variation of the duration of G-code instructions. To protection IP in AM systems, we propose a novel method called movement obfuscation to inject fake movements in a printing process to confuse an attacker. In addition, we propose an optimization-based method to find optimal patterns to be injected into the side-channels to obstruct side-channel attacks.