Font Size: a A A

Research And Implementation Of Adaptive VoIP Steganography Based On Matrix Embedding

Posted on:2018-10-10Degree:MasterType:Thesis
Country:ChinaCandidate:P YueFull Text:PDF
GTID:2358330512976778Subject:Control engineering
Abstract/Summary:PDF Full Text Request
As an important branch of information hiding,steganography has become a hotspot in the field of information security.Voice over IP(VoIP)has become one of the most important carrier types for steganography because of its huge amount of users and the persistence of data flow.In recent years,adaptive VoIP steganographyhas become one of the main directions of audio steganographywith the popularization of matrix embedding framework represented by Syndrome-Trellis Codes(STC).In this paper,we study adaptive VoIP steganography from the aspects of carrier channel selection,steganographic distortion function design and matrix embedding framework optimization.The main research works are summarized as follows:(1)An adaptive VoIP steganographic scheme based on column vectors of speech frame is proposed.In typical Classifying Least Significant Bit(LSB)grouping and Frame Window(CLFW)scheme.G.723.1 speech frames are grouped into column vectors so as to reduce the influence of the speech quality caused by the modification of the LSB bits in the same group,and different rate linear codes are selected to construct variable-length matrix embedding for LSB bits according to the difference of their auditory redundancy,then the adaptive VoIP steganography can be realized.The experimental results show that the proposed scheme can effectively improve the auditory concealment when compared with CLFW.(2)An adaptive VoIP steganographic scheme based on Linear Predictions Coding(LPC)cepstrum distortion function is proposed.Based on the universal wavelet relative distortion(UNIWARD),the LPC cepstrum distortion function is designed.The average Cepstrum Distortion(MCD)is analyzed.Using the distortion function,the LPC cepstral coefficients of the speech frame are modified with STC to complete the embedding of the secret information.The results of Perceptual Evaluation of Speech Quality(PESQ)show that this scheme has a significant improvement on security when compared with some typical VoIP steganographic schemes.(3)Based on PJSIP,Ffmpeg and JRTPLIB open source library,the VoIP covert communication platform is designed.The signaling module is implemented by using Session Initiation Protocol(SIP),and the voice data is transmitted with Real-time Transport Protocol(RTP).The VoIP algorithm and G.723.1 encoder/decoder set are integrated as the covert communication module,the realized VoIP covert communication tool can be implemented in Windows platform.The test results show that this tool can be used for covert file transmission while maintaining normal VoIP communication.Finally,the dissertation is concluded and the problems need to be further studied are pointed out.
Keywords/Search Tags:VoIP, Steganography, Speech frame column vector, Matrix embedding, Distortion function
PDF Full Text Request
Related items