Abstract. Authentication and authorisation has been a basic and necessary service for internet transactions. With the evolution of e-commerce, traditional mechanisms for data security and access control are becoming outdated. Several new standards have emerged which allow dynamic access control based on exchanging user attributes. Unfortunately, while providing highly secure and flexible access mechanisms is a very demanding task, it cannot be considered a core competency for most e-commerce corporations. Therefore, a need to outsource or at least share such services with other entities arises. Authentication and Authorisation Infrastructures (AAIs) can provide such integrated federations of security services. They could, in particular, provide attributebased access control (ABAC) mechanisms and mediate customers' demand for privacy and vendors' needs for information. We propose an AAI reference model that includes ABAC functionality based on the XACML standard and lessons learned from various existing AAIs. AAIs analysed are AKENTI, CARDEA, CAS, GridShib, Liberty ID-FF, Microsoft .NET Passport, PAPI, PERMIS, Shibboleth and VOMS.
In highly open systems like the Internet, attributebased access control (ABAC) has proven its appropriateness. This is reflected in the utilization of ABAC in authentication and authorization infrastructures (AAIs). However, specification and maintenance of ABAC policies has turned out to be complex and error-prone even in federations of limited size, especially if heterogeneous attribute schemes are involved. Here, the arising Semantic Web can contribute to a solution. This paper describes an architecture for embedding the access control process into a semantic context employing external knowledge in form of ontologies. We base our proposal on extensions of established open standards. Using the approach presented, policy management at the different sites of a federation is simplified by a semantic attribute management facility.
Authentication and Authorisation Infrastructures (AAIs) are gaining momentum throughout the Internet. Solutions have been proposed for various scenarios among them academia, GRID computing, company networks, and above all eCommerce applications. Products and concepts vary in architecture, security features, target group, and usability containing different strengths and weaknesses. In addition security needs have changed in communication and business processes. Security on the internet is no longer defined as only security measures for an eCommerce provider against an untrustworthy customer but also vice versa. Consequently, privacy, data canniness, and security are demands in this area.The authors define criteria for an eCommerce provider federation using an AAI with a maximum of privacy and flexibility. The criteria is derived concentrating on b2c eCommerce applications fulfilling the demands. In addition to best practices found, XACML policies and an attribute infrastructure are deployed. Among the evaluated AAIs are Shibboleth, Microsoft Passport, the Liberty Alliance Framework, and PERMIS.
Abstract. Authentication and Authorisation Infrastructures (AAIs) support service providers on the internet to outsource security services. Motivations for their usage stem from software engineering and economics. For the latter an assessment of inherent risks is needed. In this work the authors deduct an appropriate, formalistic risk assessment method for AAIs and analyse outsource able security services in comparison to traditional -non AAI involved -service providing. To achieve the assessment of risks various methods for risk management have been analysed and finally a suitable qualitative method has been chosen. As AAIs differ in their potential to cover security services, combinations of these services are compared. The given risk assessment method enables providers to decide on a special infrastructure for their purpose and lets users of AAIs determine if given advantages surpass the immanent risks. This work also enables service providers to estimate costs for such an infrastructure and calculate potential savings.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.