Classification of Encrypted Text and Encrypted Speech (Short Communication)

  • Rajesh Asthana Scientific Analysis Group, DRDO, Delhi
  • Neelam Verma Scientific Analysis Group, DRDO, Delhi
Keywords: Feature extraction, projection pursuit techniques, minimum distance classifier, maximum likelihood classifier,


The information to be exchanged between two parties can be text data or speech data. This data is encrypted for its security and communicated (to the other end). When an adversary intercepts these encrypted data then in order to recover the actual information, his first step is to identify whether intercepted data is encrypted text or encrypted speech are used. The next step is to get the actual information from encrypted text or encrypted speech. In this paper, pattern recognition techniques are applied for identification of encrypted text and encrypted speech. Some new and modified feature extraction techniques have been used to convert the text and speech data into three-dimensional, four-dimensional, and five-dimensional measurement vectors. These multi-dimensional measurement vectors are converted into two-dimensional vectors using projection pursuit technique based on Sammon.s algorithm and Chien.s algorithm. The quantified classification performances using minimum distance classifier and maximum likelihood classifier have also been given.

Defence Science Journal, 2010, 60(4), pp.420-422, DOI:

Author Biographies

Rajesh Asthana, Scientific Analysis Group, DRDO, Delhi

Done postgraduation in Mathematics from the University of Gorakhpur. He joined Scientific Analysis Group (SAG) in 2003 and is presently working as Scientist C.

Neelam Verma, Scientific Analysis Group, DRDO, Delhi

Done her post graduation in Mathematics from IIT Delhi. She joined SAG, DRDO in 1986 and at present she is working as Scientist F.

How to Cite
Asthana, R., & Verma, N. (2010). Classification of Encrypted Text and Encrypted Speech (Short Communication). Defence Science Journal, 60(4), 420-422.