Analysis of privacy and stability to interference in the communication channel of network steganography methods

Abstract

In this paper, for the first time, the effectiveness of network steganography methods has been investigated using channel coding of data when they are transmitted over communication channels with interferences, and the method of mathematical modeling evaluates the stability of network steganography methods to the detection. The network steganography methods concealing data in such protocols as HTTP, TCP and ICMP were considered. Using the ICMP «Secure» mode of the developed method, the greatest throughput was achieved. To study the noise resistance of the methods before transmission by the communication channel, the packets were encoded using the 2 Binary 1 Quandary linear coding algorithm, after which additive white Gaussian noise was added to the stego object. The paper substantiates the possibility of using network steganography methods to protect biometric data from unauthorized access in the case of using remote biometric authentication. Based on the research results, it was determined that the network steganography method based on hiding data in TCP is more efficient when working via communication channels with noise. It makes possible to recover a message with small distortions at a dispersion value of 0.4. But according to a set of criteria for resistance to noise and detection probability, the best is the «Fast» mode of the network steganography method based on hiding data in ICMP.