A new approach for root-causing attacks on digital microfluidic devices

Recent research on security aspects of digital microfluidic biochips (DMFBs) has revealed several alarming backdoors in the established DMFB design flow that can lead to unacceptable compromises in assay results, often leading to catastrophic consequences. This has led to a growing concern about vulnerability issues of DMFBs and assay manipulation attacks that are used for clinical diagnostics in healthcare. In this paper, we propose a novel scheme for localization and root-causing of assay manipulation attacks. Our approach takes as input a golden implementation of an assay and the attacker modified one, and uses symbolic reasoning on the actuation sequences of the golden assay and the compromised one to derive possible operations that may have been compromised in the modified assay. Additionally, our method can utilize intermediate checkpoint observations to derive more precise attack locations. Experimental results show the efficacy of our proposal.