中国邮电高校学报(英文) ›› 2012, Vol. 19 ›› Issue (4): 29-33.doi: 10.1016/S1005-8885(11)60279-3

• Networks • 上一篇    下一篇

Enhanced calculation of necessary QoS for user satisfaction with a QoS mapping matrix

吴吉祥,夏靖波   

  1. 1.Institute of Information and Navigation, Air Force Engineering University 2.Institute of Radar and Electronic Warfare, Equipment Academy of Air Force
  • 收稿日期:2011-11-02 修回日期:2012-06-04 出版日期:2012-08-31 发布日期:2012-09-12
  • 通讯作者: 吴吉祥 E-mail:Jixiang19850126@163.com
  • 基金资助:

    陕西省自然科学基础研究计划项目

Enhanced calculation of necessary QoS for user satisfaction with a QoS mapping matrix

Ji-Xiang WU 2   

  1. 1.Institute of Information and Navigation, Air Force Engineering University 2.Institute of Radar and Electronic Warfare, Equipment Academy of Air Force
  • Received:2011-11-02 Revised:2012-06-04 Online:2012-08-31 Published:2012-09-12
  • Contact: Ji-Xiang WU E-mail:Jixiang19850126@163.com
  • Supported by:

    Natural Science Foundation Research Project of Shanxi Province

摘要:

The final goal of quality of service (QoS) guarantees is to assure high quality of experience (QoE) for users. We cannot control QoE, but it is feasible to use QoS control schemes at lower levels to keep QoE high. We need to know the necessary QoS parameters values which need the minimum cost but could satisfy required QoE. An enhanced method is proposed to get the necessary quality of service (QoS) parameters which cost minimum but could satisfy required Quality of Experience (QoE) well. It abstracts the problem of obtain necessary QoS parameters to be how to get the solution of linear regression equations considering cost coefficient. The method establishes equations utilizing principal component analysis and multiple regression analysis based on normalized data. This paper defines the cost of necessary QoS and introduces the calculation of minimum cost necessary QoS parameter by inverse or generalized inverse matrix operations. The numerical example proves the normalization to be effective and the calculated values are feasible.

关键词:

user satisfaction, QoS, normalization, multiple regression analysis, cost, generalized inverse

Abstract:

The final goal of quality of service (QoS) guarantees is to assure high quality of experience (QoE) for users. We cannot control QoE, but it is feasible to use QoS control schemes at lower levels to keep QoE high. We need to know the necessary QoS parameters values which need the minimum cost but could satisfy required QoE. An enhanced method is proposed to get the necessary quality of service (QoS) parameters which cost minimum but could satisfy required Quality of Experience (QoE) well. It abstracts the problem of obtain necessary QoS parameters to be how to get the solution of linear regression equations considering cost coefficient. The method establishes equations utilizing principal component analysis and multiple regression analysis based on normalized data. This paper defines the cost of necessary QoS and introduces the calculation of minimum cost necessary QoS parameter by inverse or generalized inverse matrix operations. The numerical example proves the normalization to be effective and the calculated values are feasible.

Key words:

user satisfaction, QoS, normalization, multiple regression analysis, cost, generalized inverse

中图分类号: