On Mon, Dec 28, 2015 at 2:26 AM, Colby Boyer <[email protected]> wrote:
> Hi GNURadio folks, > > Some portions of the candidate standard for ATSC 3.0 have been posted > to the ATSC website (http://atsc.org/standards/candidate-standards/), > and specifically the PHY/bootstrap sections are now available. > > As far as I know there is no one working on a GNURadio version, but I > thought it would be worth while to check for collaborators before > venturing out. > > Note, this would be a personal project for me not for a dissertation or > work. > > Thanks, > Colby > Colby, Well, no one's responded, so it looks like you might be on your own here. But keep us up to date on your progress with it. I'd definitely like to see how you're getting along with things. While we're on the subject of ATSC, I wanted to point out our current implementation available in-tree in gr-dtv. It works great but is still very CPU intensive. I've done a bit with it myself to improve a few of the blocks for speed using VOLK (I'll get these checked in soon). However, the heaviest block by far is the Viterb decoder block. Looking at how it's done, it feels to me like it's just waiting for the right VOLK kernel to be built to take car of a lot of the processing. If we can cut this block's CPU time in half or more, we can probably run real-time on a reasonably configured machine. So Colby, that's also a suggestion for you: any time to see a place to speed up the code through vectorization, either do or make a note of it for other developers to pay attention to in the future. Thanks! Tom
_______________________________________________ Discuss-gnuradio mailing list [email protected] https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
