The Study Of Application On Secret Communication Based On Audio Watermark Technology

With the rapid development of communication and computer technology, informationcan be shared and exchanged more easily and conveniently. However, the security issues ofcommunication are increasingly shown, and the traditional encryption-based securecommunication has been researched for many years and made lots of achievements. However,as the encrypted contents are easy to attract the attention of the third party during transmission,it has failed to meet the security needs of resolving the current issue. As an emerging field ofinformation security research in recent years, digital watermarking has gained a lot ofattention and made great development. Similar to watermarking technology, informationhiding can embed secret information in the carrier file, which uses the transparency ofembedded algorithm in public networks, so that the third party couldn’t find the existence ofthe secret communication during transmission. And this technology provides a new solutionfor secret communication. The content of this thesis mainly consists of the following twoaspects：1. Compared with the audio, research on watermarking secret information embedded inthe image is more mature, and the sensitivity of the human auditory system has made greatdifficulty to research on embedding audio files. According to the characteristics of the audiowatermark itself, this research puts forward the improved algorithm combining the wavelettransform and Fourier transform. The secret taking audio transported by the MATLABsimulating unsafe channel can resist the attack from filtering, compression, quantization, andGaussian white noise. The simulation shows that the combination method of calculationshowed superior performance in both the subjective evaluation and the quantitative analysis.Compared with the original information, the extracted secret information has a highernormalized correlation count and lower error rate. The dual transformation transparency androbustness of the algorithm is superior to the single performance when using a transformation.2. Based on the dual transformation algorithm combining wavelet transform and Fouriertransformation, this research mainly introduces the following three security enhancementtechnique and thought:(1) Traditional audio covert communication technologies onlyconsidered the data features of the carrier, but ignored the relationship between the embeddedsecret data and the carrier. The program of this research considers the goodness of fit betweenthem. When embedding the secret information, it will dynamically choose the best embeddingstarting position according to the fits. Reduce the size of carrier file modification and enhancethe robustness and transparency of the algorithm under the same embedding capacity.(2) Forthe key management of embedding and extraction algorithm, this research introduced theideas of cross-domain authentication, enhanced in open public network transmission and thesecurity of use, which made the system having a higher ability against the attack.(3) Join thetimestamp in the sender information, allows the receiver to determine whether a secret message was intercepted and tampered according to this information, which enhanced thesecurity of secret communication system.