Galois Awarded $2.7 Million Navy Contract To Develop New Cyber Resilience Capability

Project award by Office of Naval Research (ONR) will focus on protecting real-time and embedded military software systems by making them resilient to attack

Galois today announced that it has been awarded a $2.7 million contract by the Office of Naval Research (ONR) to leverage software brittleness for cyber defense.

Modern military systems involve a vast array of software, which serves as an attractive target for adversaries. Protecting this software is challenging, as it often runs on legacy hardware that lacks modern cyber defenses. Furthermore, many software defense techniques impose unacceptable overheads for embedded and real time systems, which form the core of control systems for most air, ground, and naval vehicles.

With this project, Galois aims to harden control systems by using binary rewriting to add brittleness to legacy binaries. Brittleness causes programs to fail fast when under attack, which allows systems to quickly detect and disrupt cyber attacks and revert to known-good states. Coupled with the fault tolerance mechanisms that support critical systems, brittleness adds resilience to cyber systems. Galois aims to further develop and leverage its research on brittle software as a cyber defense to protect Navy-relevant software systems against control flow integrity attacks, code injection attacks, memory corruption, and overflow vulnerabilities.

“An adversary can wreak havoc if they gain control of a sensitive software system that is operationally-sensitive,” said Dr. Tristan Ravitch, Principal Investigator on the project. “Many traditional cyber defenses strive to keep systems running when cyber attacks are detected.  Our approach aims to improve overall system resilience by triggering built-in system recovery methods as quickly as possible.  This allows the system to seamlessly restart in a known-good state while denying the attacker access to a running system.”

Each project phase will add new capabilities in the main research thrust areas: 1) new methods for adding brittleness to binaries 2) efficiency of brittle code, and 3) assurance through verification.  As a cyber defense for constrained and legacy systems, brittle software has two key properties:

• Opportunism: Brittle software is opportunistic in that it can be applied as much or as little as desired and still protect the program. Brittleness can be added to software while meeting time and space constraints, as each instance of brittleness is small and independent. Moreover, brittleness requires only local transformations and does not require a global view of the program.
• Pervasiveness: Brittleness is pervasive in that it is distributed throughout a program and is enforced as a side effect of execution. There is no external monitor to bypass, forcing attackers to overcome the brittleness of the entire system.

By the end of the project, Galois aims for the tool to be suitable for evaluation on Navy-relevant systems.  Galois will identify suitable evaluation targets in coordination with the Office of Naval Research and Naval Surface Warfare Center.

For additional information on this project, visit https://galois.com/project/brittle.

About Galois

Galois has been performing computer science research and development since 1999. With many of the world’s foremost experts in computer science and mathematics and a world-class team of programmers and engineers, Galois is uniquely positioned to take on the world’s most difficult challenges in computer science. Galois is a trusted partner in the defense and intelligence industries, proving the feasibility of cutting edge research as it applies to critical systems. Technology companies turn to Galois to build reliability, safety and security into their product development efforts from day one. For additional information, visit http://www.galois.com.

This material is based upon work supported by the Office of Naval Research under Contract No. N68335-17-C-0241.

Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the Office of Naval Research.