hann window for PAPR reduction

I'm using hann windowing for PAPR reduction. After windowing the peak
value and  average value of signal reduces and overall PAPR increases
after dividing peak value with average value.what should i do to maintain
the average value of the signal and to reduce only peak value?


---------------------------------------
Posted through http://www.DSPRelated.com

Clipping ;-)

Seriously, your posting is not clearly written.

PAPR reduction is always a tradeoff with error rate (or efficiency).
Many systems (and myself) use polar clipping (to reduce PAPR), followed
by filtering (to restore original bandwidth). That's the easiest method.

More sophisticated methods use special constellations for OFDM etc.

I have not heard about using hann windowing so far ...

Peter


On 2015-07-07 15:23, theresajc wrote:
> I'm using hann windowing for PAPR reduction. After windowing the peak
> value and  average value of signal reduces and overall PAPR increases
> after dividing peak value with average value.what should i do to maintain
> the average value of the signal and to reduce only peak value?
> 
> 
> ---------------------------------------
> Posted through http://www.DSPRelated.com
> 

Well there is an interesting point.
I was under the impression that if you clip a signal to reduce the PAR, 
And then filter it to remove the resulting  spectral regrowth, that the filtering basically puts 
The PAR back to what it was.
In other words, there is a more fundamental connection between PAR or trajectory
and BW that nature does not allow you to escape.
Comments?

Mark 

Peter Mairhofer  <63832452@gmx.net> wrote:

>Clipping ;-)
>
>Seriously, your posting is not clearly written.
>
>PAPR reduction is always a tradeoff with error rate (or efficiency).
>Many systems (and myself) use polar clipping (to reduce PAPR), followed
>by filtering (to restore original bandwidth). That's the easiest method.
>
>More sophisticated methods use special constellations for OFDM etc.
>
>I have not heard about using hann windowing so far ...

I have applied Gaussian windowing for the purpose of reducing PAPR
in OFDM.  It works quite well.  It may be provable that a Gaussian
window is optimal for this purpose, but I'll leave that to the theorists.
Maybe a Hann is better.

(The above was in situations where one does not have a feee hand
to re-design constellations.)

Steve