| Service-oriented computing is a main computing model in an open distributed environment. In order to improve the reusability of Web services, we need to functionality limited number of Web services in service composition in accordance with the service description, constraints, available resources, to achieve a combination of user-defined objectives, produces value-added services.Many research institutions home and abroad have conducted quite a few researches on dynamic Web service composition, and have made some valuable achievements. However, the computing framework, modeling theory and other key technologies are not well developed. the main problems are service-oriented applications of the new architecture; services abstraction of virtualization and real-time composition; service selection, services-matching and services replacement algorithm; service quality; how to use semantic and ontological; how efficient the service quality management to meet customer service requests, etc.The contributions of this dissertation are as follows:(1)A new proposed multi-layer structure service composition model based on computing framework.According to the user's request, service composition find a path to meet the user's requests and QoS demand, so that the various service packages for each service composition path load balancing distribution as much as possible. This thesis propose multi-layer structure of the service composition model for dynamic service selection and implementation services to replace the provision of multi-dimensional QoS support, the part of research studies dynamic service composition and provide uniform service computing framework.This paper presents an efficient multi-tier structure of service composition algorithm based on the model of services. Simulation results confirmed that the algorithm have improved in rate, combination of time, load balancing, and dynamic environment, flexibility and scalability significantly compared with the traditional.(2) A new proposed Web service composition algorithm based on particle swarm algorithm.In the Web environment, there are lots of services with different QoS parameters that can satisfy the functional requirements. How to select a collection from candidate services, assure its QoS to achieve Pareto optimal. This paper presents a Web service composition algorithm based on particle swarm algorithm. Theoretical analysis and experimental results demonstrate the feasibility and effectiveness.(3) A new proposed trust evaluation model Web services.Establish trust relationships between entities is an important prerequisite for Web service.The traditional Web service model trust is not easy to detect fraud and does not reflect the impact of the independent trust.This paper presents a new Web service trust evaluation model. The model can reduce the workload of trust management center, improve overall system performance. Meanwhile, ensure the security of the system.(4) A new proposed trust-sensitive Web services composition strategy based on black-white board.Because of the formation and disappearance of services is dynamic, different services have different service-quality. It is difficult to choose a reasonable service composition. Service participants are likely malicious, fraudulent, false possibilities thus can not guarantee the credibility of the parties involved. To solve these problems, this paper proposes a trust-sensitive services strategy based on black-white board. The strategy evolved through the blackboard and whiteboard trust model to help users get reliable high-quality services. White-board records service quality and trust evaluation information in the interaction. Black-board establishes service quality and trust evaluation system to other service entities. Theory analysis and experimental confirmation of the effectiveness of this strategy. it is easy to overcome the direct trust rare small, lackeing of pre-trust issues, effectively identify the conspiracy to deceive. |
PDF Full Text Request