Re: [Discuss-gnuradio] Minimal Hardware to Run Gnuradio on?

[Tom Huppi]

On Mon, 20 Dec 2004, John Gilmore wrote:

<snip>

> The community will get around to handheld software radio units --
> probably by repurposing some standard radio built by a major mfr.  But
> part of why we wrote GNU Radio before building any hardware was that
> understanding the signal processing, and programming it so that it can
> usefully run in realtime, is a large fraction of the work.

My suspicion is that for certain potential uses of GnuRadio,
realtime processing will be difficult/delayed at best and
impractical/impossible at worst.  For instance, a key may not be
available in real time.  It also may be useful to sample a signal
(or many) and backtrack upon discovering interesting content thus
making better use of available cycles no matter how many there
are.  Basically there is a real-time device made for just about
every signal but almost none of them have the ability to act in
non realtime whether it could be useful or not.  That unique
potential aspect of GnuRadio and similar technologies really
interests me.

For whatever reason caching and queuing has always appealed to me
and in the modest design work I've done, I've been happy to have
employed it whether it turned out to be needed or not.  If nothing
else it has helped me with debugging, but the systems I've worked
with are several worlds below the GnuRadio work.  I've seen how
the thought of non-realtime work drives a lot of people to
distraction too :)

Thanks,

 - Tom