[aprssig] Aprx v2 beta, now with Viscous Digipeater
scott at opentrac.org
Thu Oct 22 11:07:15 CDT 2009
> Having said that, a typical 8 bit microcontroller probably doesn't
> have enough RAM to implement this.
The 512 bytes in the OT1+ would be a little tight. With 2k of RAM in
the MC9S08AW32 MCU I'm using for the OT1+ RT and a few other projects it
would certainly be doable.
The T2 uses a list of 16-bit hashes for tracking dupes, with an 8-bit
time-to-live for each, so that's 3 bytes per entry. Currently the list
is 15 entries long, so 45 bytes of storage total.
To implement this on the T2 shouldn't require more than one extra bit of
storage per entry. I don't see why it's necessary to count the actual
number of times it's heard - either you get it once (which automatically
results in an entry in the dupe list) or you get it more than once, and
set the excess flag. I can reduce the TTL timer to 7 bits and not
increase the storage requirements, though it'll mean a bit more code in
flash to process the bitfield.
The T2 can already defer digi processing of a packet after the APRS
parsing, KISS output, MONITOR display, and everything else has been done
(and in fact any one of those can be deferred) - as long as this is an
on/off option and doesn't require more analysis of the path to decide
whether to act on it, it shouldn't take more than a few lines of code to
I'm tempted to go try it out now, but at the moment I've got to go fire
up the forklift and make some space in the warehouse for weather
stations that are due in today.
More information about the aprssig