[1]章乐贵,邢长友,余 航,等.GAN-TM:一种基于生成对抗网络的流量矩阵推断机制[J].计算机技术与发展,2022,32(02):51-57.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 008]
　ZHANG Le-gui,XING Chang-you,YU Hang,et al.GAN-TM:A Traffic Matrix Inference Mechanism Based onGenerative Adversarial Network[J].,2022,32(02):51-57.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 008]
点击复制

GAN-TM:一种基于生成对抗网络的流量矩阵推断机制()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
32
期数:
2022年02期
页码:
51-57
栏目:
大数据分析与挖掘
出版日期:
2022-02-10

文章信息/Info

Title:
GAN-TM:A Traffic Matrix Inference Mechanism Based onGenerative Adversarial Network
文章编号:
1673-629X(2022)02-0051-07
作者:
章乐贵; 邢长友; 余 航; 郑 鹏
陆军工程大学 指挥控制工程学院,江苏 南京 210007
Author(s):
ZHANG Le-gui; XING Chang-you; YU Hang; ZHENG Peng
School of Command and Control Engineering,Army Engineering University of PLA,Nanjing 210007,China
关键词:
数据完整性; 卷积生成对抗网络; 流量矩阵推断; 信息缺失; 数据恢复
Keywords:
data integrity; convolutional generative adversarial networks; traffic matrix inference; information loss; data restoration
分类号:
TP181
DOI:
10. 3969 / j. issn. 1673-629X. 2022. 02. 008
摘要:
随着当前互联网技术的快速发展,网络规模和复杂度不断提高,由于流量矩阵对于网络管理、流量工程、异常检测等都具有重要意义,因此准确测量流量矩阵对于计算机网络而言极其重要。 当前针对流量矩阵的测量机制主要可以分为直接测量法和估计推断法,其中估计方法又包括简单统计反演法、附加链路测量信息法以及测量反演结合法。 现有测量机制在准确性和测量耗费方面存在较多问题,直接测量的方法虽然可以保证准确性,但网络规模的扩张及网络结构的日趋复杂化使其在实现上存在困难,而流量矩阵推断问题在线性求解上固有的高度病态特性又使得估计推断法时常难以发挥作用,因此需要一种新的方法以更通用的方式解决现有问题。 该文借鉴生成对抗网络( GAN) 在图像恢复方面的作用,提出了一种基于生成对抗网络的流量矩阵推断机制 GAN-TM。 GAN-TM 能够基于部分测量信息,建立起基于掩码矩阵评估的卷积生成对抗网络模型,利用部分测量信息对缺失的流量矩阵进行推断。 实验结果表明,在数据缺失率低于 30% 的情况下,GAN-TM 的推断误差能够控制在10% 以内。
Abstract:
With the rapid development of current Internet technology,the scale and complexity of network are increasing. Because trafficmatrix is of great significance for network management,traffic engineering and anomaly detection,it is extremely important to accuratelymeasure traffic matrix for computer network. At present, the measurement mechanism of traffic matrix can be divided into directmeasurement method and estimation and inference method, and the estimation method includes simple statistical inversion method,additional link measurement information method and measurement inversion combination method. The existing measurement mechanismhas many problems in accuracy and measurement cost. Although direct measurement method can guarantee the accuracy, but theexpansion of network scale and the network structure of the increasingly complicated in the implementation difficulties, inference andtraffic matrix in the height of the inherent in solving linear pathological features and makes estimation method was often difficult to play arole,so need a new kind of method in a more general way to solve the existing problems. In this paper, GAN - TM, a traffic matrixinference mechanism based on generative adversarial networks (GANs) ,is proposed based on the function of GANs in image restoration. GAN-TM establishes a convolutional generation adversation network model based on mask matrix evaluation,and uses part of the measurement information to infer the missing traffic matrix. The experiment shows that when the data missing rate is 30% ,the inference errorof GAN-TM can be controlled within 10% .

相似文献/References:

[1]魏 艳,毛燕琴,沈苏彬.一种基于区块链的数据完整性验证解决方案[J].计算机技术与发展,2020,30(01):76.[doi:10. 3969 / j. issn. 1673-629X. 2020. 01. 014]
　WEI Yan,MAO Yan-qin,SHEN Su-bin.A Data Integrity Verification Solution Based on Blockchain[J].,2020,30(02):76.[doi:10. 3969 / j. issn. 1673-629X. 2020. 01. 014]

更新日期/Last Update: 2022-02-10