Title :
Least favorable additive noise under a divergence constraint
Author :
McKellips, Andrew L. ; Verdu, Sergio
Author_Institution :
Dept. of Electr. Eng., Princeton Univ., NJ, USA
fDate :
29 Jun-4 Jul 1997
Abstract :
An uncertainty class of additive noise probability density functions (pdfs) satisfying a constraint on divergence from a prescribed nominal is analyzed in the context of antipodal communication. From within this class, the noise pdf that maximizes detection error probability is determined for both zero-threshold and maximum-likelihood detection strategies. An optional additional constraint on SNR is also considered, and asymptotic behavior is studied for vanishing divergence tolerance
Keywords :
Gaussian channels; error statistics; interference (signal); maximum likelihood detection; noise; SNR; additive noise probability density functions; antipodal communication; asymptotic behavior; detection error probability; divergence constraint; least favorable additive noise; maximum-likelihood detection strategy; uncertainty class; vanishing divergence tolerance; zero-threshold detection strategy; Additive noise; Additive white noise; Detectors; Error probability; Gaussian noise; Jamming; Maximum likelihood detection; Noise figure; Tail; Uncertainty;
Conference_Titel :
Information Theory. 1997. Proceedings., 1997 IEEE International Symposium on
Conference_Location :
Ulm
Print_ISBN :
0-7803-3956-8
DOI :
10.1109/ISIT.1997.613470
