Designing Resilient Mission Critical Systems

Mission critical systems are characterized by their extreme availability and QoS requirements, often associated with time-critical operations in potentially compromised, adversarial environments. Such systems needed to be robust and resilient in order to handle benign anomalous behavior or intentional cyber attacks and learn from experience to improve its resistance to subsequent attacks. The success of a mission is primarily depends on the uninterrupted availability of critical servers (host systems) through secure communication providing degree of performances of different subsystems and components, called mission assurance. This paper introduces a resilient architecture that provides the capabilities by combining a local monitoring and control mechanism for system applications, with distributed coordination strategies for mission management and immunization. Our approach is inspired by the way in which natural systems survive and evolve in response to environmental changes, adversaries and peer pressures.