计算机工程与应用 ›› 2021, Vol. 57 ›› Issue (17): 106-115.DOI: 10.3778/j.issn.1002-8331.2009-0002

• 大数据与云计算 • 上一篇    下一篇

面向用户需求的动态QoS服务选择方法研究

刘晶花,台宪青,马治杰,陈大鹏   

  1. 1.中国科学院 空天信息创新研究院,北京 100190
    2.中国科学院大学 电子电气与通信工程学院,北京 100049
    3.中国科学院 空天信息创新研究院 苏州研究院,江苏 苏州 215000
    4.中国科学院 微电子研究所,北京 100000
    5.江苏物联网研究发展中心,江苏 无锡 214000
  • 出版日期:2021-09-01 发布日期:2021-08-30

Study on User Requirement-Based Dynamic QoS Service Selection Method

LIU Jinghua, TAI Xianqing, MA Zhijie, CHEN Dapeng   

  1. 1.Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100190, China
    2.School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China
    3.Suzhou Research Institute, Aerospace Information Research Institute, Chinese Academy of Sciences, Suzhou, Jiangsu 215000, China
    4.Institute of Microelectronics of the Chinese Academy of Science, Beijing 100000, China
    5.Jiangsu Research and Development Center for Internet of Things, Wuxi, Jiangsu 214000, China
  • Online:2021-09-01 Published:2021-08-30

摘要:

高效、准确地为用户选择满足其需求的软件服务一直是近年来的研究热点。服务质量(Quality of Service,QoS)是衡量软件服务性能的关键指标之一,考虑到同一服务在不同网络环境下QoS值的动态性,提出一种面向用户需求的动态QoS服务选择方法,简称URDQ方法。URDQ方法采用区间数的形式记录动态环境下候选服务的属性值范围,并基于用户需求对候选服务进行初步过滤;通过区间数模型对候选服务属性区间和用户需求区间进行相对优势度计算,将属性区间数转化为易于计算的实数;使用Skyline方法对候选服务集进行过滤,减小搜索空间;根据熵权法得到的客观权重并结合用户给定的主观权重,使用TOPSIS方法对Skyline服务集进行排序。仿真实验和对比实验验证了URDQ方法在动态网络环境下的可行性和有效性。

关键词: 服务选择, 服务质量(QoS), 用户需求, 区间数模型

Abstract:

In recent years, it has been a research hotspot to select software services that are suitable for users efficiently and accurately. Quality of Service(QoS) is one of the key indicators to measure the performance of software services. Considering the dynamic of QoS values of the same service in different network environments, a user requirement-based dynamic QoS service selection method is proposed, referred to as URDQ method. Firstly, the URDQ method records the attribute value range of candidate services in dynamic environment in the form of interval number, and preliminarily filters the candidate services based on user requirements. Secondly, it calculates the relative dominance of candidate service attribute interval and user demand interval through interval number model, and converts the interval number into real numbers that are easy to calculate. Then, it uses Skyline method to filter the candidate service set to reduce the search space. Finally, according to the objective weight obtained by entropy weight method and the subjective weight given by users, the Skyline service set is sorted by TOPSIS method. Simulation and contrast experiments verify the feasibility and effectiveness of URDQ method in dynamic network environment.

Key words: service selection, Quality of Service(QoS), user requirement, interval number model