부산대학교 도서관

Cloud computing is evolved from grid computing with a key support from the rapidly expanding virtualization technology. We argue that clouding computing is particularly suitable for supporting emergency response and management (ERM) because of some of its unique features such as rapid setup and deployment on ad hoc basis, highly flexible platforms (PaaS: Platform as a Service) and application services (SaaS: Software as a Service) with little time-space constraints. ERM is one of the seven critical national infrastructures and services mandated to protect by the 1999 US President's Executive Order (PCCIP). The paradigm of survivability and survivable network systems was a response of academia to the president's executive order. We concur that survivability should be the lifeline of any ERM, including the cloud computing supported (CCS) ERM systems. In this article, we present a research agenda that is aimed at developing a survivability-centered architecture for evolving reliable and survivable CCS-ERM systems. The research agenda suggests that biological and computational evolutions should be rich sources of biological inspirations as well as powerful optimization algorithm for designing (evolving) the ERM systems. The proposed research agenda advocates the application of three-layer survivability analysis, dynamic hybrid fault models, and extended evolutionary game theory modeling developed by Ma & Krings [Ma & Krings (2008a-e, 2011), Ma et al. (2009a), Ma (2008, 2009, 2010, 2011a,b). We use banking system survivability as an example to illustrate the proposed research agenda.