Fec

GNU Radio
1

Hi,

I’m using the digital/benchmark_*.py from the GIT repository and am
getting a very low ratio of corrupt packets, usually under 0.5%. I use
the default gmsk modulation, USRP1’s with RFX2400 and run the
benchmark with a frequency of 2.4G. Would this low ratio suggest that
the FEC is being applied somewhere along the path, or not necessarily?
I cannot find any piece of code which would prove it.

Regards,
Jakub

2

On Tue, Jun 22, 2010 at 10:15 AM, Jakub M. [email protected]
wrote:

Hi,

I’m using the digital/benchmark_*.py from the GIT repository and am
getting a very low ratio of corrupt packets, usually under 0.5%. I use
the default gmsk modulation, USRP1’s with RFX2400 and run the
benchmark with a frequency of 2.4G. Would this low ratio suggest that
the FEC is being applied somewhere along the path, or not necessarily?
I cannot find any piece of code which would prove it.

Unless you made your own modifications, there is no FEC in the benchmark
code.

Thomas

3

What’s your received signal power?
For instance, this guy (no idea if he’s right) shows simulated GMSK BER
curves with a 5e-3 BER anywhere from 8db - 14db depending on the
bandwidth-time product:
http://iaci.unq.edu.ar/materias/telecomunicaciones/archivos/infoadicional/esquemas_de_modulacion_digital.pdf
Just a thought. Since it’s corrupt packets you’re talking about and not
bits, things should be a bit different, but with any decent FEC the
packet
error rate should be much lower than that.
http://iaci.unq.edu.ar/materias/telecomunicaciones/archivos/infoadicional/esquemas_de_modulacion_digital.pdf
-William