Hi all,
I’m thinking about implementing non-contiguous OFDM with gnuradio. So I
read
the ofdm codes of gnuradio. And I find the occupied tones concept in
gnuradio. Is it designed for non-contiguous OFDM? Or it is for other
reasons?
On Sun, May 2, 2010 at 4:08 AM, jf w [email protected] wrote:
Hi all,
I’m thinking about implementing non-contiguous OFDM with gnuradio. So I
read the ofdm codes of gnuradio. And I find the occupied tones concept in
gnuradio. Is it designed for non-contiguous OFDM? Or it is for other
reasons
There is an NC-OFDM implementation here:
http://www.cs.ucsb.edu/~htzheng/papyrus/detail/download.php
I am working on the code being in CGRAN with the authors.
- George
On 5/3/2010 1:48 AM, George N. wrote:
There is an NC-OFDM implementation here:
Download
http://www.cs.ucsb.edu/~htzheng/papyrus/detail/download.phpI am working on the code being in CGRAN with the authors.
- George
George,
Since the code they built to do the NC-OFDM was based originally on the
code inside GNU Radio, doesn’t it make sense to put that back into GNU
Radio instead of CGRAN? The way that I originally wrote the code was to
allow for people to do this by manipulating the subcarrier map, so as
far as I know, the change would just add functionality to the already
existing code and not change its basic functionality.
I can see having the application being in CGRAN, but by putting it all
there seems to duplicate code and will confuse the code base.
If I’m wrong and they completely reimplemented a lot of this stuff on
their own, then that’s a different story. When I talked to them at
DySPAN, it seemed like they enhanced the GNU Radio OFDM code.
Tom
George,
If I’m wrong and they completely reimplemented a lot of this stuff on their
own, then that’s a different story. When I talked to them at DySPAN, it
seemed like they enhanced the GNU Radio OFDM code.
Hey Tom,
Sure, I of course would agree it would be best in the GNU Radio code
base
It would be the best place for it, no doubt. I just figured I’d get
it
in to CGRAN first since there is absolutely 0 overhead for the code
author
(just upload as is), allowing people to get it until it is in the code
base.
This is actually a different group than the DySPAN group. This is a
group
who wrote an NSDI paper:
- George
On 05/03/2010 08:09 AM, Tom R. wrote:
Or it is for other reasons
George,
Since the code they built to do the NC-OFDM was based originally on the
code inside GNU Radio, doesn’t it make sense to put that back into GNU
Radio instead of CGRAN? The way that I originally wrote the code was to
allow for people to do this by manipulating the subcarrier map, so as
far as I know, the change would just add functionality to the already
existing code and not change its basic functionality.I can see having the application being in CGRAN, but by putting it all
there seems to duplicate code and will confuse the code base.
From a philosophy point of view, I prefer to see stuff stay outside GNU
Radio to avoid adding additional dependencies to the core GNU Radio
build. How to address duplication in this model is something that should
be considered though.
Philip