Solve/Address the Issue of Trustworthiness in Embedded Systems

University Of California-Berkeley, Berkeley CA

Investigators

Abstract

This project investigates the design, modeling, (re)configuration, and validation of trustworthy networked embedded systems (NES) applications that support the privacy of their users, the confidentiality and integrity of the data, the availability of the provided services, while implementing the required functionality and quality of service by developing the fundamental new science of secure network embedded systems and its implications for the emerging infrastructure. NES applications must be protected against malicious attacks that exploit specific vulnerabilities and characteristics of networked embedded systems. Such critical support applications that must have trusted data include equipment and process control (avionics, veitronics, communications, and SCADA/DCS systems) and environmental monitoring (pollution and chem/bio agent detection). This project seeks the following three innovations: (1) A suite of mathematical models to support the development and validation of trustworthy NES applications, (2) high confidence middleware components to assure the adaptability and survivability of NES applications, and (3) a large-scale test-bed to validate the suitability of these models and methods in a realistic NES application environment. The work is a cooperative effort of three organizations. The University of California Berkeley models distributed hybrid and embedded systems theory and platforms and addresses issues of privacy in trusted NES. UCB is building a large-scale testbed network of around 103 embedded network devices (Motes) to explore issues of NES application trustworthiness in a realistic environment and experimentally validate how the modeling and components developed by all groups in the project can protect mission-critical NES applications from potential abuses. Vanderbilt University brings their extensive experience in modeling functional capabilities of NES applications, which yields new systems theory and high confidence composable middleware frameworks with probabilistic elements to them. SRI conducts research in modeling the trustworthy aspects of NES applications and on developing methods and tools to support model-based co-design that can enable the systematic and predictable interweaving of trustworthiness with the functional applications and middleware.