Please use this identifier to cite or link to this item: http://hdl.handle.net/10603/54380
Title: Enhanced Approaches to Improve the Performance of Secure Speech Communication for Wireless Network
Researcher: Ambika D
Guide(s): Radha V
Keywords: Speaker verification
speech water marking
scrambling
Upload Date: 20-Oct-2015
University: Avinashilingam Deemed University For Women
Completed Date: 30/11/2014
Abstract: The tremendous growth envisaged in computer hardware software and network technologies have made communication in an easy and cost effective manner Recent newlinedevelopments in the communication and transmission medium have revolutionized the method of sharing and delivering contents However it also introduces serious concerns on the protection issues Secure speech communication is a tool which through the use of sophisticated software and state of the art networking technology ensures total privacy and newlinesecurity to classified speech data during transmissionThe main aim of the research work is to design a secure system that provides protection to speech data through the use of watermarking compression and encryption over wireless communication network To develop such a system the research work first uses a speaker verification system to authenticate the persons who use the system and then uses secure algorithms for protecting the speech data during transmission The speaker verification system consists of three steps namely feature extraction speaker pruning and verification During feature extraction a combined feature vector constructed using a discrete wavelet packet transformation based multiband mel frequency cepstral coefficients and linear predictive cepstral coefficients is used Speaker pruning is performed using KMeans and KNearest Neighbour algorithm The verification is based on an homogeneous ensemble model constructed using the majority voting aggregation function Three types of classifiers were analyzed for this purpose They are backpropagation neural network support vector machine and extreme learning machine The secure system is developed by first enhancing the operations of compression encryption and watermarking which are then combined to design a joint secure system Compression is performed using a Huffman coder enhanced to reduce the number of Huffman tables used using a multiple Huffman table approach The encryption algorithm is enhanced by first classifying the speech signal into most significant and
Pagination: 
URI: http://hdl.handle.net/10603/54380
Appears in Departments:Department of Computer Science

Files in This Item:
File Description SizeFormat 
dambika_chapter1.pdfAttached File7.42 MBAdobe PDFView/Open
dambika_chapter2.pdf7.49 MBAdobe PDFView/Open
dambika_chapter3.pdf7.03 MBAdobe PDFView/Open
dambika_chapter4.pdf7.28 MBAdobe PDFView/Open
dambika_intro.pdf577.12 kBAdobe PDFView/Open


Items in Shodhganga are protected by copyright, with all rights reserved, unless otherwise indicated.

Altmetric Badge: