尺度变换复双树小波网络隐藏信道深度检测

针对隐藏信道检测问题解决中传统检测算法存在特定隐藏信道盲区,或对某类隐藏信道针对性过强而忽视其他隐藏信道的问题,提出一种基于复双树小波包变换的邻域和空域联合网络隐藏信道检测。基于复双树小波包有限冗余变换所特有的平移不变特性、尺度间不同变换系数的相关性,以及尺度相同变换系数邻域间的相关特征,并结合信号增强机制,实现对变换系数的取舍,以达到隐藏信道信号增强效果;然后与块阈值算法联合对网络时间隐藏通道特征进行提取,采用深度学习方式实现隐藏信道训练和检测;最后通过在IPCTC、TRCTC、Jitter Bug、MBCTC、FXCTC五种典型时间隐藏通道中进行实验验证,显示所提算法具有更高的特征精度和较...

Full description

Saved in:
Bibliographic Details
Published in计算机应用研究 Vol. 34; no. 1; pp. 256 - 260
Main Author 路璐 凌捷
Format Journal Article
LanguageChinese
Published 广东工业大学 计算机学院,广州,510006 2017
Subjects
复双树小波
尺度变换
平移不变
隐藏信道
隐藏信道
translation invariant
平移不变
复双树小波
scale transformation
covert channel
尺度变换
dual tree complex wavelet
Online AccessGet full text

Cover

More Information
Summary:针对隐藏信道检测问题解决中传统检测算法存在特定隐藏信道盲区,或对某类隐藏信道针对性过强而忽视其他隐藏信道的问题,提出一种基于复双树小波包变换的邻域和空域联合网络隐藏信道检测。基于复双树小波包有限冗余变换所特有的平移不变特性、尺度间不同变换系数的相关性,以及尺度相同变换系数邻域间的相关特征,并结合信号增强机制,实现对变换系数的取舍,以达到隐藏信道信号增强效果;然后与块阈值算法联合对网络时间隐藏通道特征进行提取,采用深度学习方式实现隐藏信道训练和检测;最后通过在IPCTC、TRCTC、Jitter Bug、MBCTC、FXCTC五种典型时间隐藏通道中进行实验验证,显示所提算法具有更高的特征精度和较快的计算时间。
Bibliography:51-1196/TP
For covert channel detection problem solving, traditional detection algorithms exist specific covert channel blind, or of some kind of covert channel for strong and ignore other hidden channel problems, this paper proposed a method based on dual tree complex wavelet packet transform domain and spatial domain combined network covert channel detection. Firstly, based on dual tree complex wavelet packet transform limited redundancy transform characteristic of translation invariant features, different transform coefficients of correlation between scale and scale transform coefficient between neighborhood characteristics and combined with signal enhancement mechanism, choice of transform coefficients based on scale correlation, to achieve signal enhancement effect. Secondly, with block threshold algorithm were combined in the network covert timing channel features extraction and then the depth study way covert channel training and testing. Finally, the experiments were carried out in the IPCTC, TRCTC, Ji
ISSN:1001-3695
DOI:10.3969/j.issn.1001-3695.2017.01.058