PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Signal Masking in Gaussian Channels
John A. Quinn and Christopher Williams
In: ICASSP 2008, 30 Mar - 4 April 2008, Las Vegas, USA.

Abstract

We consider the problem of modifying the noise properties of a channel in order to make the source as indecipherable as possible given the output. Applications include jamming communications, maintaining confidentiality near spoken conversations and masking noise pollution. We present results as to how this can be done efficiently, assuming that we have a Gaussian channel and a constraint on the power of the noise. We go on to consider the case in which there is a positive signal which we want to remain coherent, as well as a negative signal which we wish to confound. We also discuss the application of the theory to acoustic signals, where we consider aspects of the human auditory system.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Conference or Workshop Item (Paper)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
ID Code:4668
Deposited By:Christopher Williams
Deposited On:13 March 2009