[Ground-station] LDPC work at Burning Man Math Camp

[Michelle Thompson]

Goal is to take the LDPC code and work on porting to Verilog at Math Camp
at Burning Man.

If you're going or know someone going, let them know. Math Camp has a
library, whiteboards, lecture track, a bar, lounge, and plenty of wonderful
engineers, mathematicians, scientists, supporters, and visitors.

We're working on having an FT-8 experiment as well. The noise environment
at Burning Man at HF is one of the worst in the world. You want a place to
test your signals? That's the place.

We're leaving this Thursday, 23 August and plan to return Tuesday, 4
September.

While a lot of LDPC work (and presentations!) will be happening (if all
goes as planned*), there isn't reliable communications from Burning Man.

LDPC work folds right into a talk at TAPR DCC in mid-September, and is a
large part of the Block Party at GNU Radio Conference immediately following
DCC. That work will get summarized (in whatever form it's in!) in Madrid at
the Open Source Cubesat Workshop.

None of this would be possible without the work of Charles Brain, Jan
Schiefer, and Bob McGwier. Encouragement and advice from Mike Murphree,
Brian Butler, and many local IEEE ITA people. Nate and Neel and others from
Ettus Research and GNU Radio Foundation are responsible for providing the
opportunity for the Block Party.

I'll be relying pretty heavily on Paul KB5MU to keep the explanations in
the proper scope and check the math, and Ken Easton to review the Verilog.
There will be a lot to review so expect appeals in early September before a
hard landing at DCC.

I believe that this is the hardest part of the receiver design to open
source. It's nontrivial math. Getting it into GNU Radio and figuring out
how to make it efficient and elegant are really big ambitious goals. We are
up to it. You are needed to review and improve!

A parallel effort is to get the GPU code working in the GPU accelerator in
GNU Radio, measure the performance, and demonstrate it (like a cell phone
video). Take this job and run with it! Post to the list for help and
collaboration. This job has been on the burner for a while, but not a lot
of progress has been made. There is no reason a GPU can't be the engine of
a Trans-Ionospheric radio. There is a limitation, in that efficiencies are
lost due to the restriction that all the frames be the same format, but
it's worth it for the massive parallel speedup.

If you're intrigued, dive in, ask for help, and see what you can learn and
contribute. That's why we're here! :+)

-Michelle W5NYV

*Plans will encounter a dynamic and challenging environment, but we're
confident that progress can and will be made.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openresearch.institute/pipermail/ground-station-openresearch.institute/attachments/20180821/977e847c/attachment.html>