From: volkanizma2003 on
>On Jun 29, 3:31=A0pm, "volkanizma2003"
><ozduranvolkan(a)n_o_s_p_a_m.gmail.com> wrote:
>> =A0Hi, I have some problems regarding noise normalisation. I have
origina=
>l
>> signal and I want to add a noise.
>>
>> EbN0db=3D 0.00;
>> SNRv=3D10^(EbN0db/10);
>> global sigma;
>> sigma=3D1/sqrt(SNRv);
>>
>> b=3Drandn(1,1200);
>> gaus=3Dsigma*b;
>>
>> x=round(randn(1,1200)); % my original signal
>> y=gaus % my noise
>> z=x+ gauss % original signal + noise
>>
>> Question 1.) Do I need to normalize the noise for example multiple by
0.1
>>
>> z=x+ 0.1*gauss %will it be like this?
>>
>> is this noise`s phase is equal to my original signal or less than my
>> original signal?
>>
>> Thanks for your time.
>
>If x is complex why aren't you adding complex noise?
>
>Greg
>
Dear Greg, thanks for your time and interest. X is not a complex x is a
digital signal. it contains 0 and 1 bits. And I add analog noise signal
into my digital signal. should my noise's amplitude is equal to the my
original signal's amplitude or less than original signals amplitude. do I
need to adjust the noise by multiple with 0.1

By the way, when I use Ofdm, I use complex signals as you know. Do I need
to add random complex noise, or just a random noise. thanks volkan.
From: Greg Heath on
On Jul 1, 3:48 pm, "volkanizma2003"
<ozduranvolkan(a)n_o_s_p_a_m.gmail.com> wrote:
> >On Jun 29, 3:31=A0pm, "volkanizma2003"
> ><ozduranvolkan(a)n_o_s_p_a_m.gmail.com> wrote:
> >> =A0Hi, I have some problems regarding noise normalisation. I have
> origina=
> >l
> >> signal and I want to add a noise.
>
> >> EbN0db=3D 0.00;
> >> SNRv=3D10^(EbN0db/10);
> >> global sigma;
> >> sigma=3D1/sqrt(SNRv);
>
> >> b=3Drandn(1,1200);
> >> gaus=3Dsigma*b;
>
> >> x=round(randn(1,1200)); % my original signal
> >> y=gaus % my noise
> >> z=x+ gauss % original signal + noise
>
> >> Question 1.) Do I need to normalize the noise for example multiple by
> 0.1
>
> >> z=x+ 0.1*gauss %will it be like this?
>
> >> is this noise`s phase is equal to my original signal or less than my
> >> original signal?
>
> >> Thanks for your time.
>
> >If x is complex why aren't you adding complex noise?
>
> >Greg
>
> Dear Greg, thanks for your time and interest. X is not a complex x is a
> digital signal. it contains 0 and 1 bits.

Look again. You are using round(randn), not round(rand).

Also, why do you reference phase?


> And I add analog noise signal
> into my digital signal. should my noise's amplitude is equal to the my
> original signal's amplitude or less than original signals amplitude. do I
> need to adjust the noise by multiple with 0.1
>
> By the way, when I use Ofdm, I use complex signals as you know. Do I need
> to add random complex noise, or just a random noise. thanks volkan.- Hide quoted text -
>
> - Show quoted text -

I really don't know why you are doing this.
If you want a real-valued binary signal
why not

sig(sig>=0.5) = 1
sig(sig<0.5) = 0

Greg

From: HardySpicer on
On Jun 30, 10:40 pm, John <sampson...(a)gmail.com> wrote:
> On Jun 29, 3:31 pm, "volkanizma2003"
>
>
>
> <ozduranvolkan(a)n_o_s_p_a_m.gmail.com> wrote:
> >  Hi, I have some problems regarding noise normalisation. I have original
> > signal and I want to add a noise.
>
> > EbN0db= 0.00;
> > SNRv=10^(EbN0db/10);
> > global sigma;
> > sigma=1/sqrt(SNRv);
>
> > b=randn(1,1200);
> > gaus=sigma*b;
>
> > x= round(randn(1,1200)); % my original signal
> > y= gaus % my noise
> > z= x+ gauss % original signal + noise
>
> > Question 1.) Do I need to normalize the noise for example multiple by 0..1
>
> > z=x+ 0.1*gauss %will it be like this?
>
> > is this noise`s phase is equal to my original signal or less than my
> > original signal?
>
> > Thanks for your time.
>
> 1) You already scaled the noise when you multiplied it by 1/sqrt(SNR)
> 2) Adding the noise does not change the phase, only the amplitude
>
> John

Actually it does.
First  |  Prev  | 
Pages: 1 2
Prev: Convolution property of DFT
Next: IEEE Xplore question