能耗感知的Web服务组合技术研究

发布时间：2018-11-15 08:00

【摘要】：近几年来，随着Web服务技术的发展，服务组合已被工业界、学术界普遍认为是Web服务的重要支撑技术之一。与此同时，由于Web服务的广泛使用，其领域内也出现了规模较大的组合服务。因此，大规模组合服务的能耗也相应迅速增加，这极大地影响了服务组合的整体性能。为了更好地发挥服务组合的作用，提高系统效率，以能耗优化为目标的服务组合研究刻不容缓。传统的能耗优化理论并不完全适用于服务组合领域，需要对其作进一步的拓展和改进。目前在服务组合领域内涉及能耗优化的研究内容十分匮乏，仍有很多问题急待解决。在绿色计算的大背景下，本研究以服务组合运行中的能耗优化为目标，在满足用户全局QoS约束的前提下，实现能耗感知的服务组合。在当前的服务计算环境下，降低服务组合过程中的能耗不仅可以提高服务组合的效率，同时也提高了系统资源的利用率，以此形成一个高效、节能的服务组合机制。本文从能耗的角度出发，首先对单个服务进行能耗建模。接着根据服务组合的特点推导出组合服务的能耗模型，该模型适用于面对同一个用户请求存在多个可行路径的情况。然后基于这一能耗模型，给出了非功能属性QoS约束，并得出基于状态空间模型的对应约束方程。完成整个建模过程后，，继续对该服务组合能耗模型进行了计算求解，同时也给出了能耗比、QoS约束两个计算条件。计算过程分为两步，分别是可行路径的查找和请求速率的分配。考虑到前者过程较为复杂，后文将单独列出一章来讲述具体的查找算法；而后者则直接利用混沌理论来计算，以服务组合的总能耗为目标函数，请求速率为自变量来求解。在可行路径的查找部分，文章给出了一个基于规划图的服务组合方法，利用规划图的分层概念来求解。算法过程分为三步，分别是：分层搜索阶段、最优QoS值计算阶段以及可行路径选取阶段。文章最后通过仿真实验对所提出的方法进行了正确性与有效性的验证。实验表明，与传统方法相比，本文提出的能耗感知的服务组合方法在用户全局QoS为约束的基础上，不仅能保证用户需求的满足，同时也有效减少了组合服务的总能耗。
[Abstract]:In recent years, with the development of Web services technology, service composition has been widely regarded as one of the important supporting technologies of Web services in industry and academia. At the same time, due to the wide use of Web services, there are also larger composite services in its field. Therefore, the energy consumption of large-scale composite services increases rapidly, which greatly affects the overall performance of service composition. In order to play a better role of service composition and improve system efficiency, it is urgent to study service composition with the goal of energy consumption optimization. The traditional energy consumption optimization theory is not fully applicable to the field of service composition, and needs to be further expanded and improved. At present, the research on energy consumption optimization in the field of service composition is very scarce, and there are still many problems to be solved urgently. Under the background of green computing, this study aims at optimizing the energy consumption in the service composition operation, and realizes the energy-consumption aware service composition under the premise of satisfying the user's global QoS constraints. In the current service computing environment, reducing the energy consumption in the process of service composition can not only improve the efficiency of service composition, but also improve the utilization of system resources, so as to form an efficient and energy-saving service composition mechanism. From the point of view of energy consumption, this paper firstly models the energy consumption of a single service. Then, according to the characteristics of service composition, the energy consumption model of composite services is derived. The model is suitable for the case where there are multiple feasible paths in the face of the same user request. Then, based on the energy consumption model, the nonfunctional attribute QoS constraints are given, and the corresponding constraint equations based on the state space model are obtained. After the whole modeling process is completed, the energy consumption model of the service composition is solved, and two calculation conditions, the ratio of energy consumption and the QoS constraint, are also given. The calculation process is divided into two steps, which are the search of feasible path and the allocation of request rate. Considering the complexity of the former process, a separate chapter will be listed later to describe the specific search algorithm, while the latter is directly calculated by using chaos theory, taking the total energy consumption of the service composition as the objective function and the request rate as the independent variable. In the part of finding feasible paths, this paper presents a method of service composition based on planning graph, which is solved by using the concept of layering of planning graph. The algorithm is divided into three steps: hierarchical search, optimal QoS calculation and feasible path selection. Finally, the correctness and validity of the proposed method are verified by simulation experiments. The experiments show that compared with the traditional method, the proposed energy-aware service composition method can not only guarantee the satisfaction of the user's requirements, but also reduce the total energy consumption of the composite services effectively, based on the global QoS constraints of the users.
【学位授予单位】：杭州电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09

【参考文献】