Information content in Gaussian noise: optimal compression rates

Abstract

We approach the theoretical problem of compressing a signal dominated by Gaussian noise. We present expressions for the compression ratio which can be reached, under the light of Shannon's noiseless coding theorem, for a linearly quantized stochastic Gaussian signal (noise). The compression ratio decreases logarithmically with the amplitude of the frequency spectrum $P(f)$ of the noise. Entropy values and compression rates are shown to depend on the shape of this power spectrum, given different normalizations. The cases of white noise (w.n.), $f^{n_p}$ power-law noise ---including $1/f$ noise---, (w.n.$+1/f$) noise, and piecewise (w.n.+$1/f|$ w.n.$+1/f^2$) noise are discussed, while quantitative behaviours and useful approximations are provided.

28 LateX pages and 6 Fig, replaced with minor changes to match published version

References in corpus (1)

• An Efficient Technique to Determine the Power Spectrum from Cosmic Microwave Background Sky Maps

Cited by in corpus (3)

• Noise and bias in square-root compression schemes
• The effect of signal digitisation in CMB experiments
• Data Processing and Compression of Cosmic Microwave Background Anisotropies on Board the PLANCK Satellite