discuss-gnuradio
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] Softbits in gr-ieee 802.11 ?


From: Michael Dickens
Subject: Re: [Discuss-gnuradio] Softbits in gr-ieee 802.11 ?
Date: Sat, 20 Jan 2018 12:06:56 -0500

Hi Sumit - OK good I'll take a look when I get a chance. In the meantime, 3 
thoughts come to mind:

1) If done correctly, you don't need to modify GR at all. Your new 
constellation class is purely in your OOT (or, gr-ieee-802.11 in this case). 
You create the base class in your OOT & inherit from it in your OOT too. All is 
done in the OOT; easier that way, and you can add special methods as you see 
fit, and/or fix issues.

2) In order for SWIG to do it's magic on a non-block base class & inheriting 
class, you need some special SWIG sauce. This might already be done in 
gr-ieee-802.11; I'll check when I get there. GR does it via a special file 
that's included into one of the main gr-digital .i SWIG files; for specifics 
see this file < 
https://github.com/gnuradio/gnuradio/blob/master/gr-digital/swig/constellation.i
 > & the top-level SWIG file for gr-digital here (at the very end) < 
https://github.com/gnuradio/gnuradio/blob/master/gr-digital/swig/digital_swig0.i
 >.

3) There is already a method for handling soft decoding that you can overload 
if you want to: "calc_soft_dec" < 
https://github.com/gnuradio/gnuradio/blob/master/gr-digital/include/gnuradio/digital/constellation.h#L163
 >. You might want to check it out instead of creating a new method.

If you succeed over the weekend, please do let us know. If I get some spare 
time to look into this over the weekend I'll let you know. Cheers! - MLD

On Fri, Jan 19, 2018, at 5:30 PM, sumit kumar wrote:
> Yes I just made one and pushed everything. 
> 
> https://github.com/sumitstop/MTSDR-gnuradio/commits/master
> 
> https://github.com/sumitstop/MTSDR-gr-ieee-80211/commits/master
> 
> My changes are in latest commit.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]