Gaussian Noise
Understanding Gaussian Noise
The Gaussian noise model is the foundation of all receiver and communication system analysis. It captures the statistical properties of thermal noise, which is the dominant noise source in most RF systems. Understanding Gaussian noise is essential for calculating sensitivity, SNR, and BER.
Properties
- White: Equal power per Hz across all frequencies. Power spectral density = kT (W/Hz).
- Gaussian: Amplitude follows a Gaussian (normal) distribution. Most samples are near zero; large amplitudes are rare but not impossible.
- Additive: Noise adds to the signal without multiplying or distorting it.
- Independent: Noise samples at different times are statistically independent.
AWGN Channel
The AWGN channel is the simplest communication channel model: received signal = transmitted signal + Gaussian noise. It represents the best-case scenario (no fading, no multipath, just noise). All modulation schemes are first characterized in AWGN.
Frequently Asked Questions
What is AWGN?
AWGN (Additive White Gaussian Noise) is random noise with flat spectral density and Gaussian amplitude distribution. It is the standard noise model for communication systems, representing the thermal noise present in all receivers. It sets the fundamental performance limit.
Why is the Gaussian assumption important?
The Gaussian (normal) distribution accurately describes thermal noise and allows closed-form calculation of BER for many modulation schemes. The central limit theorem guarantees that the sum of many independent noise sources approaches a Gaussian distribution.
Is real noise always Gaussian?
Thermal noise is very accurately Gaussian. However, real environments also contain non-Gaussian noise: impulse noise (switching, lightning), man-made interference, and quantization noise. System designs should account for non-Gaussian components where they are significant.