| Accelerating the development of modern service industry is the basic idea of the National Twelfth Five-Year Plan. Developing the operational common services could effectively lower the threshold of company business development, so that the companies only need to focus on their own core business and improve their core competitive power. Reusable common services could also be standardized to avoid "islands of information" problem and help the companies promote efficiency. However, the existing common service delivery platform has some technical flaws, such as service delivery mode, service selection mechanism, service selection performance and so on. So the Qos couldn’t be guaranteed when delivering the common service, and the idea of common service can’t be realize.In order to effectively address the QoS problems of the common services, we need new technical-driven force to help bring the evolution of common service platform. According to the distributed computing and cloud computing technology, this paper proposed a QoS-driven network. The developing direction of modern service network can be the architecture evolution and the node function evolution.1) The first is network architecture evolution towards distributed network. Because of the performance limitation of the centralized business provision mode, we consider the distributed computing technology, transferring the single node ability provision problem into multiple nodes cooperation problem, which will be derived from a single service delivery platform for the unification of the homogenization by the network nodes to provide and manage hosted services.2) The second is the evolution of node function. Besides introducing the distributed computing technology to reconstruct the common services provision mode, we still need to improve the business performance of single node. The single node performance is the basic factor of the whole service network performance. So we consider the virtualization, to transfer the ability delivery problem of single node into resource delivery problem. Through unified resource management system and optimized business capability of deploy and management, Improve the resource utilization of single homogeneous business nodes, so that to improve the QoS.The main innovation of this paper is:1. Proposed a new system of modern service network architecture and service node functionality, high reliability and QoS guarantee as a starting point, the derivative of a unified service delivery platform for hosting network services, service delivery and management:service network consists of homogenization of service nodes, service nodes distributed through the introduction of technology service platform to provide centralized, and by optimizing the mix of services and routing mechanism for efficient combination of services, improve business QoS; Taking into account of the large when the scale of operations of each service node server’s physical resources, there is a large-scale patterns, the introduction of virtualization technology, providing the physical resources of loosely coupled, service node by the service layer and resource layer to improve the effective integration of services to improve business efficiency QoS.(Chapter2)2. Service Composition provides aflexible framework for new service construction by the integrating atomic services developed independently. Algorithms are needed to select atomic service and service instances with the various QoS levels which according to some application-dependent performance requirements. Our objective of service selection is to maximize a QoS function under the end-to-end QoS constraints, as well as reducing the complexity and computing cost. In this Chapter the problem is modeled as a multi-dimension multi-choice0-1knapsack problem. A novel efficient heuristic algorithm for service selection is presented. This algorithm takes the priority of service components and the priority of QoS attributes. These priorities are useful in promoting the efficiency and reducing the complexity of service selection algorithm. The simulation results show that the improvement on efficiency is obvious and increasing with the number of service components.(Chapter3)3. Proposed a new decentralized service routing mechanism, using a distributed approach to avoid the concentration of DHT routing services to the shortcomings of the signaling plane and media plane of the processing of the service node through the homogenization process, the effective use the performance of each node to improve business efficiency in routing and execution services to improve business QoS. In this Chapter, taking the advantage of load balancing in a P2P scheme, we propose a novel approach to support efficient service composition using a peer-to-peer (P2P) service routing algorithm using aggregated route. Evaluation results shows that our design for service routing outperforms traditional P2P algorithms in a service oriented network.(Chapter4)4. Proposed a new service-oriented virtual resource management system (STRATUS), reconstruction of the homogenization of the system service node based platform, through the introduction of virtualization technology to provide the physical resources of loosely coupled. Based virtual machine hosting service node based on features, to provide policy-based management system to optimize service deployment, service monitoring and service life cycle scheduling in order to achieve the level of resources sufficient for the perception of the upper layer services, in order to effectively improve service provide a platform for performance and service capabilities of QoS. STRATUS of the proposed system to complete the design and development, the system has to support projects in the country,"a safe and reliable reproductive health services, telecom operation support system for key technologies" subject "modern population and reproductive health of public service platform" has been application.(Chapter5)... |
PDF Full Text Request