[RFC] libosmocore: New libosmosdr sub library
Hi, I've just pushed a libosmosdr sublibrary in a sylvain/sdr branch in the libosmocore git. Currently it contains mostly a complex vector 'class' along with various math functions used on them (convolution / correlation / ...). Some aspects the the API along with some of the underlying math are inspired by the code from openbts, but it is a complete rewrite in pure C with some aspects simplified, and some others extended. If anybody has comments before I merge it in master, they're welcome. The doxygen doc should be clear enough (and even includes fancy latex equation :P) Cheers, Sylvain
Hi Sylvain, On Mon, Oct 10, 2011 at 09:22:19AM +0200, Sylvain Munaut wrote:
I've just pushed a libosmosdr sublibrary in a sylvain/sdr branch in the libosmocore git.
is there any compelling reason to include it in the libosmocore repository? I would consider it as substantially independent of libosmocore and rather put it into an extra repository. I know we don't have a very clean situation with libosmo{core/gsm/vty} in one repository, but I think at least new libraries should go into separate repositories. What do you think? If you agree, I'll create a repository. Regards, Harald -- - Harald Welte <laforge@gnumonks.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)
Hi Sylvain, On Fri, Oct 14, 2011 at 02:00:11PM +0200, Sylvain Munaut wrote:
What do you think? If you agree, I'll create a repository.
Sure, sounds fine to me.
done: libosmo-sdr.git -- - Harald Welte <laforge@gnumonks.org> http://laforge.gnumonks.org/ ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6)
On Mon, Oct 10, 2011 at 3:22 AM, Sylvain Munaut <246tnt@gmail.com> wrote:
Hi,
I've just pushed a libosmosdr sublibrary in a sylvain/sdr branch in the libosmocore git.
Currently it contains mostly a complex vector 'class' along with various math functions used on them (convolution / correlation / ...). Some aspects the the API along with some of the underlying math are inspired by the code from openbts, but it is a complete rewrite in pure C with some aspects simplified, and some others extended.
This is great. I previously did an experimental C rewrite for convolution in openbts to work with NEON optimizations. https://github.com/ttsou/openbts-uhd/commit/b333c476 When I work on signal processing aspects again - primarily for the e100 - I'll base it off your library. Thomas
participants (3)
-
Harald Welte -
Sylvain Munaut -
Thomas Tsou