TABLE OF CONTENTS As far as the laws of mathematics refer to reality, they are not certain; and as far as they are certain, they do not refer to reality
(A. Einstein)
Electrical noise theory has the reputation of being arcane and rather complicated. There are several reasons for this:
To understand the physics of noise, as well as noise modeling and computation, background information is needed from various fields like signal theory, circuit theory, quantum mechanics, wave propagation, probability, and statistical analysis.
As a matter of fact, the phenomena which are pertinent to noise generation are usually investigated with statistical methods and/or thermodynamic concepts. To describe random processes, we use the concepts of probability, average, and correlation. Then, we must take into account that circuits process noise exactly like any useful signal; therefore, we use the Fourier series and integral approach and power or energy spectra, as well as the concept of transfer function.
If we look at the location of noise sources in any electronic circuit, we realize that they are distributed everywhere. For computational convenience and supposing that the circuit can be regarded as a two-port, we prefer to lump the noise into two equivalent voltage or current generators, conveniently located at the circuit terminals. This approach involves the...
