基于信息熵理论的多元回归网络性能评价  

Multiple Regression Network Performance Evaluation Based on Information Entropy Theory

作　　者：燕晓波 熊伟晴 刘华[1] 程子桃[1] 

机构地区：[1]中国电子科技集团公司第二十七研究所,郑州450047

出　　处：《电光系统》2015年第4期28-33,共6页Electronic and Electro-optical Systems

摘　　要：文章提出了基于信息熵理论的面向应用多元回归网络性能评价模型。该评价模型从面向应用和服务的角度，将定性与定量的评价方法结合起来，给出了网络性能的四个级别，提出了网络性能评价值和累积网络性能评价值两个新的网络性能评价指标，并给出了它们的多元回归数学模型。在该评价模型中，采用客观赋权法中的熵权法计算多元线性回归模型中各个评价属性的权重。实验结果表明，本文提出的评价模型可以对网络系统平台的性能进行可靠准确可信的评价。最后，对实验结果进行了分析讨论评价，通过定性评价结果与定量评价结果的一致性对比，再次表明了本文提出的评价模型的准确性和可靠性。This paper addresses a network performance evaluation model of application-oriented multiple regres- sion based on information entropy theory. The evaluation model, from the perspective of application and serv- ice, combines qualitative evaluation method and quantitative evaluation method. Then, it gives four levels of network performance, proposes network performance evaluation value and cumulative network performance eval- uation value, and their multiple regression mathematical models are given. The model calculates the weight of each attribute with entropy method. Experimental results show that the proposed model can provide reliable, ac-curate and credible network performance evaluation. Finally, the experimental results are analyzed and evaluated. The consistent comparison between qualitative evaluation results and quantitative evaluation result shows again the reliability, accuracy and credibility of the proposed network performance evaluation model in this paper.

关 键 词：网络性能评价 信息熵理论 多元回归 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]