Coding along Hermite polynomials for Gaussian noise channels

  • Authors:
  • Emmanuel A. Abbe;Lizhong Zheng

  • Affiliations:
  • IPG, EPFL, Lausanne;LIDS, MIT, Cambridge, MA

  • Venue:
  • ISIT'09 Proceedings of the 2009 IEEE international conference on Symposium on Information Theory - Volume 3
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper shows that the capacity achieving input distribution for a fading Gaussian broadcast channel is not Gaussian in general. The construction of non-Gaussian distributions that strictly outperform Gaussian ones, for certain characterized fading distributions, is provided. The ability of analyzing non-Gaussian input distributions with closed form expressions is made possible in a local setting. It is shown that there exists a specific coordinate system, based on Hermite polynomials, which parametrizes Gaussian neighborhoods and which is particularly suitable to study the entropic operators encountered with Gaussian noise.