# Power and variance of Gaussian White Noise

Started by February 26, 2004
```hi there

this is a newbie working on a signal denoising problem. i want to jam
a given signal with a white noise pulsed jammer such that the
effective jammer to signal ratio is some known value, say 30 db for
example.

to generate the wgn, i specify the variance in simulink for the
gaussian noise source.

what is the realtion between the variance (sigma) and power of white
gaussian noise?

could anyone guide to links to formulae for power for other
distributions?

regards, nityanand
```
```Arjun Admirer wrote:
> hi there
>
> this is a newbie working on a signal denoising problem. i want to jam
> a given signal with a white noise pulsed jammer such that the
> effective jammer to signal ratio is some known value, say 30 db for
> example.
>
> to generate the wgn, i specify the variance in simulink for the
> gaussian noise source.
>
> what is the realtion between the variance (sigma) and power of white
> gaussian noise?

I think you mean sigma^2 here.

>
> could anyone guide to links to formulae for power for other
> distributions?
>
> regards, nityanand

Try this:

Gaussian_sigma = 0.5 * sqrt(2) * ( 10^(0.05 * noise_level_dB ) )

Good luck,
OUP
```
```In article 98a8c1a8.0402261015.41c8dc5d@posting.google.com, Arjun Admirer at
nityanand_misra@hotmail.com wrote on 02/26/2004 13:15:

>
> what is the realtion between the variance (sigma) and power of white
> gaussian noise?

the variance of a random process is the AC power.  the mean square is the
total, AC+DC, power.  but the power of *true* white noise is infinite.
that's because there is infinite area under the power spectrum for white
noise.  bandlimited white noise has finite power that is proportional to the
bandwidth.

> could anyone guide to links to formulae for power for other
> distributions?

same as above.

r b-j

```