网络包长度隐蔽信道的建模与仿真

Modeling and Simulation of Covert Channel Based on Network Packet Length

隐蔽信道既是网络中进行隐蔽通信的重要工具,也是黑客窃取信息的手段,因而对它的研究具有重要的意义。提出了将网络包长度作为载体进行隐蔽通信的模型。为了评估隐秘信道,提出了信道容量、抗检测性、鲁棒性及安全性等评估指标。由于隐蔽长度信道发送数据时不受网络包发送时延的限制,隐蔽长度信道的容量远大于传统隐蔽时间信道的容量;且该信道受网络结构、网络流量的影响较小,因而具有较好的鲁棒性。通过仿真实验可知,包长度隐蔽信道的抗检测性、安全性都比传统的时间信道好。因此,隐蔽长度信道克服了传统隐蔽信道不稳定、易被干扰的缺点,是一种可靠、稳定的隐蔽通信方式。另外,使用的基于神经元网络的隐蔽信道检测法也可有效地解决隐蔽长度信道带来的危害。

Research of covert channel is of great significance not only because it is an important tool of network steganography,but also it is a frequently used mean to steal information by hacker.Packet length covert channel was proposed,hiding information with the length of network packet.And channel capacity,robustness,safety and undetectability were proposed to assess a covert channel.Because of avoidance of hiding message by the delay of sending packet,the capacity of packet length covert channel is much larger than traditional covert time channel.Moreover,the changing of network structure and the quality of network may make the covert time channel out of work.But these factors cannot play a role in packet length covert channel,so that the robust of packet length covert channel is better than covert time channel.Hence,covert channel by packet length is a more stability and useful tool of covert communication.The detection algorithms by packet length based on neuron net could work well to eliminate the risk of covert channel.