Those algorithms for Speech Enhancement may be able to fix your problem.
And if the noise is stationary, the simplest spectral substraction with a
simple VAD will be OK.

>Dear Colleagues,
>
>I have a problem. I have velocity signals x(t) recorded with an acoustic
>Doppler velocimeter. These signals are contaminated with a white noise
>n(t), so x(t) = d(t) + n(t), where d(t) is a true velocity signal.
>
>I can know the noise variance, noise power spectrum Gnn(f) and the true
>power spectrum as 
>Gdd(f) = Gxx(f) � Gnn(f).
>
>I am looking for a digital filter which removes the white noise from my
>signals in the time domain, Is there a filter for that? I wish to obtain
>d(t).
>
>I am using MatLab.
>
>Thanks for your attention
>
>Sincerely 
>
>Mart�n.
>
>
>
>
>

"Mat&#4294967295;nArg" <mromagnoli@ina.gov.ar> wrote in message 
news:__ydnfOux9Jg6_XZnZ2dnUVZ_tidnZ2d@giganews.com...
> Dear Colleagues,
>
> I have a problem. I have velocity signals x(t) recorded with an acoustic
> Doppler velocimeter. These signals are contaminated with a white noise
> n(t), so x(t) = d(t) + n(t), where d(t) is a true velocity signal.
>
> I can know the noise variance, noise power spectrum Gnn(f) and the true
> power spectrum as
> Gdd(f) = Gxx(f) - Gnn(f).
>
> I am looking for a digital filter which removes the white noise from my
> signals in the time domain, Is there a filter for that? I wish to obtain
> d(t).
>
> I am using MatLab.
>
> Thanks for your attention
>
> Sincerely
>
> Mart&#4294967295;n.
>

Hello Martin,

Look into Kalman filtering. It works quite well for this kind of thing.

Clay

Mat=EDnArg wrote:
> Dear Colleagues,
>
> I have a problem. I have velocity signals x(t) recorded with an acoustic
> Doppler velocimeter. These signals are contaminated with a white noise
> n(t), so x(t) =3D d(t) + n(t), where d(t) is a true velocity signal.
>
> I can know the noise variance, noise power spectrum Gnn(f) and the true
> power spectrum as
> Gdd(f) =3D Gxx(f) - Gnn(f).
>
> I am looking for a digital filter which removes the white noise from my
> signals in the time domain, Is there a filter for that? I wish to obtain
> d(t).
>
> I am using MatLab.
>
> Thanks for your attention
>
> Sincerely
>
> Mart=EDn.

clues only for homework problems...

what is the bandwidth of your desired signal?

what is the bandwidth of the undesired white noise signal?

Mark

Dear Colleagues,

I have a problem. I have velocity signals x(t) recorded with an acoustic
Doppler velocimeter. These signals are contaminated with a white noise
n(t), so x(t) = d(t) + n(t), where d(t) is a true velocity signal.

I can know the noise variance, noise power spectrum Gnn(f) and the true
power spectrum as 
Gdd(f) = Gxx(f) � Gnn(f).

I am looking for a digital filter which removes the white noise from my
signals in the time domain, Is there a filter for that? I wish to obtain
d(t).

I am using MatLab.

Thanks for your attention

Sincerely 

Mart�n.