It is well understood that the increasing importance of open distributed systems and the growing number of services that utilise them has given rise to the need for effective distributed systems management . Furthermore, it has also become increasingly difficult to build management systems that can cope with the increasing size and complexity of current open distributed systems [192, 196].
Recently, a number of distributed object-based computing environments, such as the Common Object Request Broker Architecture (CORBA)  and the Telecommunication Intelligent Network Architecture (TINA) , have been developed that aim to support interoperability between domains in large open distributed systems, where components may appear and disappear dynamically at run-time. As a direct result of the interoperability promoted by such architectures, the complexity of security and management policies has increased greatly due in part to the requirement to support both mutual transparency and individual autonomy between domains , and also because of the necessity to cope with the coexistence of multiple, potentially inconsistent, security and management policies.
In this thesis, an approach to the modelling of policy and a formal temporal logic based specification of a policy language has been developed that are suitable for supporting the requirements of large open distributed systems. Use of both the concepts defined in the policy model and a formal policy language has enabled the precise specification of what constitutes conflict in policy-based management systems. Furthermore, the development of methods and algorithms that support the systematic dynamic detection and resolution of policy-based conflicts in large open distributed environments have been defined.