Binary code Randomization for Attack Sensitive Software (BRASS)
Navy SBIR FY2015.2


Sol No.: Navy SBIR FY2015.2
Topic No.: N152-120
Topic Title: Binary code Randomization for Attack Sensitive Software (BRASS)
Proposal No.: N152-120-0677
Firm: Intelligent Automation, Inc.
15400 Calhoun Drive
Suite 190
Rockville, Maryland 20855
Contact: Gregory Briskin
Phone: (301) 294-4755
Web Site: http://www.i-a-i.com
Abstract: Specific mission requirements exist for certain types of critical cyber systems that demand for the software to fail/crash, as soon as possible, when under cyber-attack, rather than to operate in degraded or compromised state. In order to achieve this objective, BRASS offers novel binary code randomization approach that addresses these requirements in two phases: (1) a static phase that transforms a target binary into randomizable form, and (2) a load-time phase that randomizes the binary by reordering its instructions each time it starts. BRASS applies these techniques in order to achieve �fast-crash� property of guaranteed prompt execution termination in case of successful ROP-based exploitation attacks that cause compromised software to operate in unsafe state. BRASS does not require a source code, and can be applied to a legacy software already deployed by DoD. BRASS provides a proof of functional equivalence for all software variants generated using BRASS binary transformation method. BRASS guarantees accomplishes acceptable size overhead and no degradation in system performance. Each generated variant maintains the same level of robustness as the original program under normal condition. BRASS provides the level of diversification sufficient to guarantee that the attack that succeeded in a single variant, will crash other variants.
Benefits: BRASS can be directly applied to the military and intelligence contexts. Both government and commercial organizations will benefit from the development of BRASS that can be used as a binary code transformation tool for software with a long life time; in a situations where source code and development infrastructure is no longer available or not economically feasible; in the environment where mission critical software infrastructure is under perpetual APT threat and software integrity and confidentiality is more important than resiliency, e.g. operating in degraded and potentially compromised state is not allowed. The DoD, along with several other government organizations (e.g., DHS, DoE, states, etc.) are actively pursuing multiple cyber security related technology development programs These factors provide significant motivation for DoD and government adoption of BRASS. The extensibility of BRASS framework to support embedded software systems would allow for easy transition to specific Navy, such as Hall Mechanic Engine (HME) embedded control processors and controllers. BRASS has significant commercial potentials as a binary code transformation framework that can be applied to mission critical software systems operating in a hostile Internet environment. BRASS targets the market segment, whereas compromised or misbehaving software could be much more dangerous than simple unavailability, and, also whereas redundant and diversified backup systems are readily available. A faster failure and timely switch-over would minimize the disruption/damage and actually enhance overall resiliency.

Return