White noise provides a good approximation to real signals input to linear systems. The white noise approximation greatly simplifies the computation of the output of a linear system. EE 5500 Probability and Stochastic Processes, 2014 by J. B. Burl
Definition
A white noise signal has:
Zero mean. Correlation function is zero at all times other than zero. RX ( ) = S X ( )
The PSD of a white noise signal is a
constant: = S X ( ) F= [ RX ( )] S X
EE 5500 Probability and Stochastic
Processes, 2014 by J. B. Burl
Comments
The impulse appearing in the definition of white
noise means that the variance and total power of white noise is infinite (one reason why this signal does not exist in nature).
Note that putting white noise into a filter with
finite bandwidth yields a well defined variance and power for the output. White noise will provide a useful approximation to signals input to filters. We will see when this approximation is good after analyzing the output of systems with white noise inputs. EE 5500 Probability and Stochastic Processes, 2014 by J. B. Burl