Computer systems today are large, complex, and distributed, supporting applications that are critical to life, health, and safety. Such systems are vulnerable to attacks and intrusions. Survivable systems are systems that continue to operate correctly and reliably despite intrusions that cause some nodes to behave in an arbitrary or malicious manner, and despite faults or accidents that affect nodes or communication between nodes.
This project will advance scientific knowledge about survivable distributed systems by (1) developing a general framework for reasoning about, characterizing, designing, and evaluating survivable distributed systems, and (2) designing and implementating a survivable system architecture that is scalable and can adapt to changes in the environment. In the first effort, ways to reason about survivability and to model trustworthy survivable systems will be developed. These will extend fundamental understanding of characteristics and qualities desirable in survivable systems. In the second thrust, a new survivable system architecture will be designed. This system will consist of three regions, each of which provides differing levels of security, reliability, and quality of service. This project will take place at a national liberal arts college and involve undergraduate students in survivable distributed systems research, preparing them for further work in the design, implementation, and deployment of such systems.
