Research On Speech Content Authentication In Parameter Domain

With the development and popularity of network technology, speech communication has been vigorously developing. Modern audio editing and processing techniques reduce the cost of high-quality forgery. In order to achieve security of audio applications, a security verification system is required to ensure the security of speech content.Once part or whole of the content of the speech was tampered, it may cause some serious consequences such as the cases that businessmen pay extra money and the armies execute tampered orders, etc. Therefore, a speech content authentication method is needed to authenticate whether the speech content is tampered or not. Different from the traditional technologies like digital signature or digital watermarking, speech content authentication algorithm must tolerate the common content preserving operations such as recompression, transcoding, re-sampling and so on. The algorithm also can detect whether the speech content was tampered and can indicate the tamper location at the same time.Firstly,in this thesis the related theories of speech content authentication are introduced, existing speech coding standards are compared with each other and implement speech content authentication integrated with the speech coding that has better integrative capability. A parameter domain speech content verification algorithm with special speech coding is designed, the method of verification coefficients selection, calculating and authentication data comparing are explained. Against the content preserving operations and speech content tamper in the transmission of speech coding, experiments of the proposed algorithm and the existing algorithm are done, and the results are compared.Secondly, the security issues in authentication data transmission are considered. A transmission method that uses cryptography and error correction code to ensure the security of authentication data is investigated.Experimental results demonstrate that with the selected speech coding, compared to existing algorithm, the proposed algorithm has better ability to distinguish content preserving operations from content tamper, and it can implement the tamper localization more accurately. Therefore, the proposed algorithm can be used to solve the security issues in commercial communications, confidential telephone and military order transmission.