Multiband FM transmission reception problem

Hello,I have modified the fmradio.py andfmtx4.py code to make a multi
band fm transmission system.The problem is as I increse the number of
channels,The interference increases.How do i control the gain and also
the SNR of the transmisison.Regards, Mayur Sarode Dear discussgnuradio!
Get Yourself a cool, short @in.com Email ID now!

ANyone can please help?

mayur_CEN wrote:


View this message in context:
http://www.nabble.com/multiband-FM-transmission-reception-problem-tp23498950p23525789.html
Sent from the GnuRadio mailing list archive at Nabble.com.

On Wed, May 13, 2009 at 09:51:43AM -0700, mayur_CEN wrote:

Asking 3 times in one morning is unlikely to get you what you want.

When people fail to get a response to a posting, it’s generally
because they’ve failed to follow the suggestions described here:
http://gnuradio.org/trac/wiki/ReportingErrors

Eric

Please don’t spam the list, the people who know the most about
gnuradio and the usrp, for the most part, have other full time jobs
and provide help on this list as a courtesy. A response time of
several days is not unreasonable in such situations.

With that in mind, the reason you may not get a response is because
your question is rather vague, there are endless reasons for noise in
a system, not mention we have no idea what modifications you made to
stock files. Please see the guidelines for questions here:

http://gnuradio.org/trac/wiki/ReportingErrors

In this case we probably need access to a copy of your modified code
(please do not copy and paste into this message) and a more detailed
explanation of what you mean when you say ‘the interference
increases’, how are you measuring the interference.

Though I can’t say I fully understand the gnuradio/USRP combo, I can
say that accurately measuring SNR of the transmission is a difficult
thing to do; that is, there is no direct numerical conversion between
transmit power in gnuradio and the power output from the antennas.
Please see search the mailing list for discussions on this, I found
one right away for bpsk: Gaining RSSI, SNR, CNR or BER information - GNU Radio - Ruby-Forum

I hope you get your application working soon,

Jason

I am sorry for being so impatient.Actually I have been experimenting
with
this problem for a long time,so i am a bit eager to find the
solution.You
guys are doing a great job helping out people with their problems.I did
not
mean to offend anyone.

I have uploaded the code.I will explain to you in detail how i am
conducting
the experiment.
I am using 2 daughter boards for sending and receiving simultaneously.
I am checking the output of the fm transmission by running a nbfm_tx
receiver program simulated in grc
When I send 1 channel ,the audio can be heard clearly
As I increase the number of channels,the audio strength decreases which
I
think is due to interference.
What I want to know is how do I increase the signal strength and reduce
noise as I increase the number of channels.

Thanks for all your help
Mayur


View this message in context:
http://www.nabble.com/multiband-FM-transmission-reception-problem-tp23498950p23526649.html
Sent from the GnuRadio mailing list archive at Nabble.com.

mayur_CEN wrote:

When I send 1 channel ,the audio can be heard clearly
As I increase the number of channels,the audio strength decreases which I
think is due to interference.
What I want to know is how do I increase the signal strength and reduce
noise as I increase the number of channels.

What you are seeing as noise is most likely clipping.

As you increase the number of transmitted signals, you need to decrease
the power per signal since you need to stay maintain linearity.

For example, say you are transmitting 1 signal at max amplitude and
power. When you move to 2 signals, you need to use 1/2 the max
amplitude for each signal, which will give you 1/4 the max power in each
signal. If you have X signals, you need to give each one 1/X the max
amplitude, which will give each signal 1/X^2 of the max power.

If X is very big (more than 20 or so), you don’t necessarily have to
reduce the amplitude quite as much as 1/X, but for small X you do.

Matt