基于拍卖理论的Web服务选择研究与实现

发布时间：2018-02-22 03:05

本文关键词： Web服务服务组合逆向Vickrey拍卖服务质量　出处：《扬州大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着网络技术和电子商务的迅猛发展,功能单一的Web服务往往无法满足人们日益复杂的业务需求。如何有效集成分布于网络中的各类服务,实现具备强大功能的组合服务越来越受到人们的关注。但是,由于Internet的规模增大以及网络环境的动态性,不同的服务提供商往往会为了最大化自身利益虚报其服务的QoS,因此,传统的Web服务组合方法已无法在有效时间内找到一组满足用户需求的最优Web服务组合。针对这一问题,本文提出了一种基于Vickrey拍卖理论的Web服务组合方法,该方法一方面确保Web服务提供者不会虚报其服务的QoS,另一方面还考虑组合Web服务的全局QoS约束和组合流程中的单一节点的局部QoS约束,主要工作如下： 1.提出了一个基于拍卖理论的Web服务组合模型该模型主要包括：抽象的服务组合框架(Auction-Based Service Composition Framework, ABCF)和QoS组合模型 ABCF由三个参与者、两个功能模块和一个服务注册中心组成。参与者分别是User、Broker和Seller。User向Broker发送组合服务请求；Broker负责服务的选择,并将结果返回给User: Seller提供具有相应QoS约束的Web服务。功能模块包括拍卖模块和全局优化模块。拍卖模块负责Web服务的拍卖和局部服务选择；全局优化模块负责对QoS约束调整。服务注册中心提供服务发布和查询功能。 User在选择服务时不仅关注服务的功能性需求,也关注服务QoS属性。由于不同类型QoS评价标准不同,无法直接通过他们大小来确定Web服务的好坏。因此,本文提出了QoS组合模型,首先将各QoS规范化到同一维度,然后采用加权求和的方法计算每个服务的效用值,最后给出组合服务QoS计算方法。 2.提出两个基于逆向Vickrey拍卖的服务选择算法在上述拍卖模型的基础上,本文提出了一个基于逆向Vickrey拍卖的服务选择算法(RVBSSA)。算法包括两个阶段：局部选择阶段和全局调整阶段。在局部选择阶段,通过逆向Vickrey拍卖为组合流程中的每个抽象服务节点选择效用值最大的前K(K0)个具体服务。全局调整阶段首先判断选局部选择阶段选择的一组最优具体服务是否满足全局QoS约束,若满足则该组服务即为最优解,否则按照用户的偏好高低对不满足用户约束的QoS属性进行降序排序,然后利用逆向Vickrey拍卖策略依次调整相应的QoS值。此外,本文在RVBSSA算法的基础上,提出了基于逆向Vickrey拍卖协商的RVBSSA2算法,通过拍卖协商激励Seller提供更加真实的QoS属性值。通过将这两个算法与整数规划(IP)算法比较分析,论文的算法消耗时间远小于IP算法,且性能接近IP算法。 3.原型系统的设计与实现根据ABCF框架,利用NetBeans工具模拟服务组合的过程,实现了各模块功能,用MySql实现服务注册中心中Web服务的存储。通过原型系统的运行,表明该系统可以快速找出满足用户需求的近似最优组合Web服务。
[Abstract]:With the rapid development of network technology and electronic commerce, the single Web service is often unable to meet people's increasingly complex business needs. However, due to the increasing scale of Internet and the dynamic nature of the network environment, different service providers often falsely report the QoS of their services in order to maximize their own interests, therefore, because of the increase of the scale of Internet and the dynamic nature of the network environment, different service providers often falsely report the QoS of their services in order to maximize their own interests. The traditional Web service composition method can not find a set of optimal Web service composition in the effective time. In order to solve this problem, a Web service composition method based on Vickrey auction theory is proposed in this paper. On the one hand, this method ensures that the Web service provider does not falsely report the QoS of its service, on the other hand, it also considers the global QoS constraint of the Web service and the local QoS constraint of a single node in the composition process. The main work is as follows:. 1. A Web service composition model based on auction theory is proposed. This model mainly includes: abstract service composition framework Auction-Based Service Composition Framework (ABCFF) and QoS composition model. ABCF consists of three participants, two functional modules and a service registry. The participants are user broker and Seller.User who send a composite service request to Broker to select the service. The result is returned to user: Seller to provide Web service with corresponding QoS constraints. The function module includes auction module and global optimization module. The auction module is responsible for the auction of Web service and the selection of local service. The global optimization module is responsible for adjusting the QoS constraints. The service registry provides service publishing and query functions. User not only pays attention to the functional requirements of the service, but also the QoS attribute of the service. Because different types of QoS evaluation standards are different, it is impossible to determine the quality of the Web service directly by their size. Therefore, this paper proposes a QoS composition model. First, each QoS is normalized to the same dimension, then the utility value of each service is calculated by weighted summation method. Finally, the calculation method of composite service QoS is given. 2. Two service selection algorithms based on reverse Vickrey auction are proposed. Based on the above auction model, a service selection algorithm based on reverse Vickrey auction is proposed in this paper. The algorithm consists of two stages: local selection and global adjustment. The reverse Vickrey auction is used to select the pre-KK0 (pre-KK0) with the largest utility value for each abstract service node in the composition process. The global adjustment phase first determines whether the optimal set of specific services selected in the local selection phase satisfies the global QoS constraints. If satisfied, the group of services is the optimal solution. Otherwise, according to the preference of the user, the QoS attributes that do not satisfy the user constraints are sorted in descending order, and the corresponding QoS values are adjusted in turn by using the reverse Vickrey auction strategy. In this paper, based on the RVBSSA algorithm, a RVBSSA2 algorithm based on reverse Vickrey auction negotiation is proposed, which provides a more real QoS attribute value through the auction negotiation. The two algorithms are compared with the integer programming (IP) algorithm. The algorithm in this paper consumes much less time than IP algorithm, and its performance is close to that of IP algorithm. 3. Design and implementation of prototype system. According to ABCF framework, using NetBeans tools to simulate the process of service composition, the function of each module is realized, and the storage of Web services in service registry is realized by MySql. The results show that the proposed system can quickly find the approximate optimal composition Web services to meet the needs of users.
【学位授予单位】：扬州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09

【参考文献】