Research On Audio Steganography And Audio Steganalysis

With the rapid development of multimedia technology and networking, digital information can be transferred to all over the world quickly and conveniently. However, the information security problems that accompany by are becoming world ones which should be urgently to solve in the information society. Information hiding technology is a new all-around subject born to solve this problem, and it has high security since it protects information by hiding the existence of information. As the branches of the information hiding technology, steganography and steganalysis are the most important parts of the information warfare in the era of internet. Digital steganography is used to exchange secret messages by military units and, intelligence agencies, to keep private communications and sensitive data by companies or individuals, and even to hide organizing terrorism attack by terrorists. Steganalysis is the critical technique to prevent steganography from being used illegally, and it can hold back enemies'steg-communication, prevent the loss of confidential data, and monitor illegal information and so on. Steganography and steganalysis are significant to national security and social stabilization.Steganography and steganalysis which use audio as carriers were studied in this Thesis. Two steganographic algorithms and a special steganalysis algorithm were proposed based on the research of related theory. The military application of the steganography and the steganalysis and some problems for further research were also discussed. The main works and conclusions in this thesis are as follows.①Because of the limited steganography communication research, further verification and studies of Steganography and steganalysis at parameters estimation and relative methods were made in this paper. Firstly, theory model and information hiding model were given. Then, information hiding model which takes the whole process of embedding, attacking and extracting into account was analyzed, and invisibility and robustness, etc. were further analyzed. Lastly, in view of information hiding model, the definition of capacity of information hiding was given, the quantificational relationship of imperceptibility, robustness and hiding rate in digital watermarking was also given based on it, which pointed out that the key to reach the tradeoffs of the three is the embedding intensity.②To get a better balance between robustness and imperceptibility for audio digital watermark, a steganographic algorithm based on the audio sample dots inversion and BP network was designed. The algorithms ensures the imperceptibility by inversing the audio sample dots that have little influence on the human hearing, and realizes the embedding of confidential data in the inversing way Then the corresponding relationship between original audio signals and secret audio signals is built by the learning of artificial neural network which has the ability of realizing non-linear mapping, so that the watermarked information could be better extracted. Experimental results show that the watermarked audio can resist different kinds of attacks, such as low-pass filtering, noise, Mp3 compression, echo, re-sampling. The disadvantage is relatively small capacity.③An audio steganographic algorithm in DCT and DWT domains, which had large capacity to resist the MP3 compression efficiently, was proposed. In this algorithm, the low-frequency band, which has little influence on the human hearing, was chose to be the embedding part. Then, the divided DTC transformation on the small wavelet coefficients of the chose band was made and the watermark was embedded in every band's direct current coefficient of DCT, so that the robustness of the algorithm was improved to the greatest extent on the promise of ensuring the imperceptibility. And experimental results verify that it had good imperceptibility, larger embedding capacity and strong robustness. This algorithm is significant to realize the steganographic communication by using the MP3 files in the internet. Disadvantages can not resist synchronized and DA/AD at the same time.④An novel Audio Information Hiding Algorithm Resisting DA/AD Conversions is proposed. Based on extensive experiments, we find that the similarity coefficient of the original audio's waveform and the attacked audio's waveform is high when the frame length is appropriate. And it has little influence on the human hearing after inversing the audio sample dots in a period of time. Based on those features, at first we divide the host audio into several frames, then inverse the audio sample dots to embed secret information. We extract the secret information by judging the polarity of the similarity coefficient, and the resynchronization is designed to increase the correct rate. Experimental results show that the watermarked audio has good imperceptibility and is robust against different kinds of attacks, such as noise adding, low-pass filtering, resampling, echo, Mp3 compression, especially it can resist the attacks of the DA/AD conversions and desynchronization.⑤Considering steganalysis spread spectrum steganography, a new audio steganalysis algorithm based on the wavelet denoising and wavelet discontinuity detection technologies was proposed. In this algorithm, the correct detection rate is determined by the strength of the secret message and has nothing to do with the capacity. Furthermore, the algorithm also can calculate the length of the PN sequence if the audio is a stego-audio so that it can estimate the average capacity that hidden in audio. The experimental results proved these advantages above and showed the outstanding detection performance.