Search

Steganalysis seeks to detect the presence of secret data embedded in cover objects, and there is an imminent demand to detect hidden messages in streaming media. This paper shows how a new steganalysis algorithm based on Fast Fourier Transform (FFT) can be used to detect the existence of secret data embedded in streaming media. The proposed algorithm uses machine parameter characteristics and a network sniffer to determine whether the Internet traffic contains streaming channels. The detected streaming data is then transferred from the time domain to the frequency domain through FFT. The distributions of power spectra in the frequency domain between original VoIP streams and stego VoIP streams are compared in turn using t-test, achieving the p-value of 7.5686E-176 which is below the threshold. Results indicate that the proposed FFT-based steganalysis algorithm is effective in detecting the secret data embedded in VoIP streaming media.

Streaming technologies such as VoIP are widely embedded into commercial and industrial applications, so it is imperative to address data security issues before the problems get really serious. This thesis describes a theoretical and experimental investigation of secure covert communications over streaming media using dynamic steganography. A covert VoIP communications system was developed in C++ to enable the implementation of the work being carried out.\ud A new information theoretical model of secure covert communications over streaming media was constructed to depict the security scenarios in streaming media-based steganographic systems with passive attacks. The model involves a stochastic process that models an information source for covert VoIP communications and the theory of hypothesis testing that analyses the adversary‘s detection performance.\ud The potential of hardware-based true random key generation and chaotic interval selection for innovative applications in covert VoIP communications was explored. Using the read time stamp counter of CPU as an entropy source was designed to generate true random numbers as secret keys for streaming media steganography. A novel interval selection algorithm was devised to choose randomly data embedding locations in VoIP streams using random sequences generated from achaotic process.\ud A dynamic key updating and transmission based steganographic algorithm that includes a one-way cryptographical accumulator integrated into dynamic key exchange for covert VoIP communications, was devised to provide secure key exchange for covert communications over streaming media. The discrete logarithm problem in mathematics and steganalysis using t-test revealed the algorithm has the advantage of being the most solid method of key distribution over a public channel.\ud The effectiveness of the new steganographic algorithm for covert communications over streaming media was examined by means of security analysis, steganalysis using non parameter Mann-Whitney-Wilcoxon statistical testing, and performance and robustness measurements. The algorithm achieved the average data embedding rate of 800 bps, comparable to other related algorithms. The results indicated that the algorithm has no or little impact on real-time VoIP communications in terms of speech quality (< 5% change in PESQ with hidden data), signal distortion (6% change in SNR after steganography) and imperceptibility, and it is more secure and effective in addressing the security problems than other related algorithms.