Biblio

In this paper, we consider the problem of decentralized verification for large-scale cascade interconnections of linear subsystems such that dissipativity properties of the overall system are guaranteed with minimum knowledge of the dynamics. In order to achieve compositionality, we distribute the verification process among the individual subsystems, which utilize limited information received locally from their immediate neighbors. Furthermore, to obviate the need for full knowledge of the subsystem parameters, each decentralized verification rule employs a model-free learning structure; a reinforcement learning algorithm that allows for online evaluation of the appropriate storage function that can be used to verify dissipativity of the system up to that point. Finally, we show how the interconnection can be extended by adding learning-enabled subsystems while ensuring dissipativity.

Hardware vulnerabilities are often due to design mistakes because the designer does not sufficiently consider potential security vulnerabilities at the design stage. As a result, various security solutions have been developed to protect ICs, among which the language-based hardware security verification serves as a promising solution. The verification process will be performed while compiling the HDL of the design. However, similar to other formal verification methods, the language-based approach also suffers from scalability issue. Furthermore, existing solutions either lead to hardware overhead or are not designed for vulnerable or malicious logic detection. To alleviate these challenges, we propose a new language based framework, QIF-Verilog, to evaluate the trustworthiness of a hardware system at register transfer level (RTL). This framework introduces a quantified information flow (QIF) model and extends Verilog type systems to provide more expressiveness in presenting security rules; QIF is capable of checking the security rules given by the hardware designer. Secrets are labeled by the new type and then parsed to data flow, to which a QIF model will be applied. To demonstrate our approach, we design a compiler for QIF-Verilog and perform vulnerability analysis on benchmarks from Trust-Hub and OpenCore. We show that Trojans or design faults that leak information from circuit outputs can be detected automatically, and that our method evaluates the security of the design correctly.