|dc.description.abstract||An enhanced observer is model that observes behaviour of a service and then automatically reports any changes in the state of the service to evaluator model. The e-observer observes the state of a service to determine whether it conforms to and obeys its intended behaviour or policy rules. E-observer techniques address most problems, govern and provide a proven solution that is re-usable in a similar context. This leads to an organisation and formalisation policy which is the engine of the e-observer model. Policies are used to refer to specific security rules for particular systems. They are derived from the goals of management that describe the desired behaviour of distributed heterogeneous systems and networks. These policies should be defended by security which has become a coherent and crucial issue. Security aims to protect these policies whenever possible. It is the first line of protection for resources or assets against events such as loss of availability, unauthorised access or modiﬁcation of data. The techniques devised to protect information from intruders are general purpose in nature and, therefore, cannot directly enforce security that has no universal definition, the high degree of assurance of security properties of systems used in security-critical areas, such as business, education and financial, is usually achieved by verification.
In addition, security policies express the protection requirements of a system in a precise and unambiguous form. They describe the requirements and mechanisms for securing the resources and assets between the sharing parties of a business transaction.
However, Service-Oriented Computing (SOC) is a new paradigm of computing that considers "services" as fundamental elements for developing applications/solutions. SOC has many advantages that support IT to improve and increase its capabilities. SOC allows flexibility to be integrated into application development. This allows services to be provided in a highly distributed manner by Web services. Many organisations and enterprises have undertaken developments using SOC. Web services (WSs) are examples of SOC. WSs have become more powerful and sophisticated in recent years and are being used successfully for inter-operable solutions across various networks. The main benefit of web services is that they use machine-to-machine interaction.
This leads initially to explore the "Quality" aspect of the services. Quality of Service (QoS) describes many techniques that prioritise one type of traffic or programme that operates across a network connection. Hence, QoS has rules to determine which requests have priority and uses these rules in order to specify their priority to real-time communications. In addition, these rules can be sophisticated and expressed as policies that constrain the behaviour of these services. The rules (policies) should be addressed and enforced by the security mechanism. Moreover, in SOC and in particular web services, services are black boxes where behaviour may be completely determined by its interaction with other services under confederation system.
Therefore, we propose the design and implementation of the “behaviour of services,” which is constrained by QoS policies. We formulate and implement novel techniques for web service policy-based QoS, which leads to the development of a framework for observing services. These services interact with each other by verifying them in a formal and systematic manner. This framework can be used to specify security policies in a succinct and unambiguous manner; thus, we developed a set of rules that can be applied inductively to verify the set of traces generated by the specification of our model’s policy. These rules could be also used for verifying the functionality of the system.
In order to demonstrate the protection features of information system that is able to specify and concisely describe a set of traces generated, we subsequently consider the design and management of Ponder policy language to express QoS and its associated based on criteria, such as, security. An algorithm was composed for analysing the observations that are constrained by policies, and then a prototype system for demonstrating the observation architecture within the education sector. Finally, an enforcement system was used to successfully deploy the prototype’s infrastructure over Web services in order to define an optimisation model that would capture efficiency requirements.
Therefore, our assumption is, tracing and observing the communication between services and then takes the decision based on their behaviour and history. Hence, the big issue here is how do we ensure that some given security requirements are satisfied and enforced? The scenario here is under confederation system and based on the following:
System’s components are Web-services.
These components are black boxes and designed/built by various vendors.
Topology is highly changeable.
Consequently, the main issues are:
• The proposal, design and development of a prototype of observation system that manages security policy and its associated aspects by evaluating the outcome results via the evaluator model.
• Taming the design complexity of the observation system by leaving considerable degrees of freedom for their structure and behaviour and by bestowing upon them certain characteristics, and to learn and adapt with respect to dynamically changing environments.||en