Research On Retrieval Method For Encrypted Speech Based On Content Privacy-Preserving In Cloud Environmemt

With the in-depth research and application of cloud technology,the multimedia data owners usually choose to store their large-scale data in cloud server.While enjoying the convenience of cloud storage,the privacy and security of multimedia speech data need to be protected by users.The research of speech encryption and encrypted speech retrieval technology is increasingly appreciated by relevant experts.Therefore,processing and retrieving large-scale encrypted speech efficiently has become a hot area of research.Aiming at the high efficiency of massive encrypted speech retrieval in cloud computing environment,and to ensure the security,integrity and shareability of speech data,and to solve the problem of data privacy protection and copy threat which faced by owners;this thesis researches on the key technologies of speech encryption algorithm,extracting the perceptual features from the encryption speech,and encrypted speech retrieval based on content privacy protection.The detailed contents of the research are as follows:1.In order to extract the perceptual features from the encrypted speech,a speech scrambling encryption algorithm based on discrete cosine transform was proposed.The experimental results show that the proposed algorithm can extract the perceptual features from the encrypted speech as a search digest.Furthermore,to guarantee the privacy and security of speech data stored in the cloud,a speech chaotic encryption algorithm based on Folded-towel discrete chaotic was proposed.The experimental results show that the proposed algorithm has good encryption performance and can meet the basic requirements of speech storage in cloud server.2.To solve the issue of the search digest that cannot be extracted directly from the encrypted speech via the content-based encrypted speech retrieval method,a retrieval method of encrypted speech based on short-term cross-correlation and perceptual hashing was proposed.Firstly,the speech scrambling encryption algorithm based on discrete cosine transform is used to encrypt the speech file and upload the encrypted speech data to the encryption speech database in cloud.Then,the sample speech clips are obtained by the cutting operation from the speech file for scrambling encryption.The perceptual hashing sequence of the encrypted speech is constructed by extracting the short-term cross-correlation of the encrypted speech signals as the search digest.These perceptual hashing sequences are uploaded into the hashing index table of cloud system.At the phase of searching,the Hamming distance algorithm is used for the matching retrieval operation during the search.The experimental results show that the proposed encryption speech hashing algorithm has a better discrimination,robustness and compactness.In the whole retrieval process,the downloading and decrypting operations of speech data are not necessary.3.In order to solve the problems of the existing encrypted speech retrieval scheme function are not complete,the encryption database sharing cannot be realized,and the copy threat and privacy protection issues.A retrieval scheme of encrypted speech based on perceptual hashing and watermark protocol in storage environment was proposed.Firstly,the speech owner extracts the retrieval digest by constructing an efficient speech perceptual hashing scheme.Then,local speech data is encrypted by utilizing the Folded-towel discrete chaos system,and the obtained search digest and encrypted speech are uploaded in the cloud server.In addition,,in order to prevent the illegal distribution of speech by dishonest users in cloud,a distribution protocol based on digital watermarking was proposed.At the phase of searching,the Hamming distance algorithm is used for the matching retrieval operation during the search.The experimental results show that the perceptual hashing algorithm constructed by this scheme has a better discrimination,robustness and compactness.The proposed scheme can effectively protect the privacy of owners and track the users who illegally distribute the speech.Meanwhile,it can also realize single-user to multi-user security retrieval of encryption speech,and in the whole retrieval process,the downloading and decrypting operations of speech data are not necessary.