尺度变换复双树小波网络隐藏信道深度检测

发布时间：2018-01-05 13:38

本文关键词：尺度变换复双树小波网络隐藏信道深度检测　出处：《计算机应用研究》2017年01期 　论文类型：期刊论文

【摘要】：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。
[Abstract]:Aiming at solving the problem of detecting hidden channel in the traditional detection algorithm has a certain hidden channel blind, or of a hidden channel for too strong and ignore other hidden channel problems, put forward a kind of neighborhood and spatial network dual tree complex wavelet packet transform based on hidden channel detection. Shift invariant characteristic of complex wavelet packet Co. based on the correlation of different redundant transform, transform coefficients between scales, and the same characteristic scale transform coefficients combined with neighborhood, signal enhancement mechanism, realize the transform coefficients in order to achieve the trade-off, hidden channel signal enhancement effect; and then combined to block threshold algorithm of network time hidden channel features are extracted by deep learning the way to achieve hidden channel training and detection; finally in IPCTC, TRCTC, Jitter, Bug, MBCTC, five kinds of typical FXCTC channel in real time. The verification shows that the proposed algorithm has higher feature precision and faster computing time.

【作者单位】：广东工业大学计算机学院;
【基金】：广东省科技计划项目(2013B040401017,2014A010103029,2015B010108002) 广州市科技计划项目(2014J4100201,201508010026) 广东工业大学计算机学院教学改革项目
【分类号】：TN911.7
【正文快照】： 0引言Lampson等人在1973年提出隐藏信道概念,指明隐藏信道是违背开发者意愿而被附加设计出的通信信道[1]。文献[2]首次把隐藏信道定义进行网络化扩展。隐藏网络信道是APT攻击的主要信息传输方式,对网路安全构成危害。网路内,按照信息传递载体差异,可把隐藏信道划分成时间以及

【相似文献】