运行环境敏感的信息服务动态选择与组合方法

发布时间：2018-03-04 19:54

本文选题：运行环境敏感　切入点：动态服务选择　出处：《东南大学》2015年博士论文　论文类型：学位论文

【摘要】：以计算机和互联网为代表的当代信息技术发展迅猛,信息以前所未有的速度增长。服务计算较好地解决了分布式网络环境下信息资源的利用问题,为满足用户需求提供了高质量的解决方案。在互联网环境下,存在着大量的服务资源,实现信息服务的有效选择与组合成为当前的重要研究课题。当前,信息服务所依赖的运行环境日益复杂和多样,给服务选择与组合带来了诸多挑战。一方面,在动态且难以预知的运行环境中,服务QoS动态变化,会导致选择的服务质量下降,甚至服务选择失败,大大降低了用户体验；另一方面,随着绿色计算的兴起,计算系统的能耗问题受到越来越多的关注,而运行环境与服务能耗密切相关,动态的运行环境可能会在服务组合过程中引发较高的服务能耗,造成能源的巨大浪费。传统研究工作的重点大多在于解决服务选择和组合方法中面向服务QoS优化的问题,较少考虑服务运行环境对服务选择与组合的影响。本论文首先考虑了运行环境相关的QoS建模方法,为了在动态环境中精简候选服务,提出了服务评估方法,在此基础上分别以优化QoS和优化能耗为目标,在感知运行环境状态的基础上提出了动态服务选择方法与多路径服务组合方法,具体包括以下几个方面的工作。首先,建立运行环境敏感的QoS模型。对服务运行环境进行描述,建立服务运行环境状态模型,提出运行环境敏感的QoS模型构建方法。其次,提出基于QoS历史记录的信息服务评估方法。采用服务QoS历史记录,根据优化模型对服务进行初步过滤,然后根据混合整数规划模型计算各服务的评估值,从而有效评估服务满足用户需求的能力,为服务选择和组合提供筛选依据。第三,提出QoS感知的动态服务选择方法。在前面两方面工作的基础上,以动态环境下优化服务QoS为目标,在组合服务设计阶段为各任务生成候选服务队列,在服务执行阶段依据当前运行环境进行动态选择。第四,提出能耗感知的多路径信息服务组合方法。采用了与运行环境中负载相关的服务基本能耗模型,针对连续到达的用户请求,提出以优化组合服务总能耗为目标的多路径服务组合模型,设计能耗感知的服务组合优化求解算法,该算法发现可行服务组合路径,并根据能耗模型合理分配请求流量。本论文考虑运行环境对服务选择与组合的影响,提出了相关解决方案。仿真实验以及原型系统的验证结果表明,本论文所提出的方法能够较好地适应动态的运行环境,解决动态运行环境下服务选择与组合的有关优化问题。本论文工作为构建高效的基于信息服务的应用系统提供了重要参考价值。
[Abstract]:Modern information technology, represented by computers and the Internet, is developing rapidly, and information is growing at an unprecedented speed. Service computing has solved the problem of using information resources in the distributed network environment. In the Internet environment, there are a large number of service resources, the effective selection and composition of information services has become an important research topic. The running environment on which information services depend is becoming more and more complex and diverse, which brings many challenges to service selection and composition. On the one hand, in the dynamic and unpredictable running environment, the dynamic change of service QoS will lead to the decline of service quality. On the other hand, with the rise of green computing, more and more attention is paid to the energy consumption of computing system, and the running environment is closely related to the energy consumption of service. The dynamic running environment may lead to high service energy consumption in the process of service composition, resulting in a huge waste of energy. The traditional research focuses on solving the problem of service-oriented QoS optimization in service selection and composition methods. The influence of service running environment on service selection and composition is less considered in this paper. Firstly, the QoS modeling method related to running environment is considered in this paper. In order to simplify candidate services in dynamic environment, a service evaluation method is proposed. On this basis, aiming at optimizing QoS and optimizing energy consumption, a dynamic service selection method and a multi-path service composition method are proposed based on the perception of the state of the running environment, including the following aspects of work. The QoS model of runtime environment is established, the service running environment is described, the state model of service running environment is established, and the method of constructing QoS model is put forward. This paper presents an information service evaluation method based on QoS history record. Using the service QoS history record, the service is preliminarily filtered according to the optimization model, and then the evaluation value of each service is calculated according to the mixed integer programming model. It can effectively evaluate the ability of service to meet the needs of users, and provide screening basis for service selection and composition. Thirdly, a dynamic service selection method based on QoS perception is proposed. Aiming at optimizing service QoS in dynamic environment, candidate service queue is generated for each task in composition service design stage, and dynamic selection is made according to current running environment in service execution stage. 4th, A multipath information service composition method based on energy consumption awareness is proposed, which adopts the service basic energy consumption model related to the load in the running environment, and aims at the continuous arrival of user requests. A multi-path service composition model with the goal of optimizing the total energy consumption of composite services is proposed to design an energy-aware service composition optimization algorithm. The algorithm finds feasible service composition paths. According to the energy consumption model, the request flow is allocated reasonably. In this paper, considering the influence of running environment on service selection and composition, the related solutions are proposed. The simulation results and the verification results of the prototype system show that, The method proposed in this paper can better adapt to the dynamic running environment. To solve the optimization problem of service selection and composition in dynamic running environment, this paper provides an important reference value for the construction of an efficient application system based on information services.
【学位授予单位】：东南大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：TP393.09

【相似文献】