On 23/09/16 23:50, amsat-bb-request(a)amsat.org wrote:
> Packet BBS operation via space was proved to be entirely inoperable,
> impossible total congestion,, inefficient, impractical, wasteful and
> discunctional use of packet via satllites 20 years ago.
Yes, connected AX.25 is the wrong protocol for a PBBS on the ISS, and
CSMA simply WILL NOT WORK when most user nodes can't hear each other
(aka Hidden Transmitter Syndrome).
My understanding of PBBS operation is that the vast majority of users
"download" far more data than they "upload", and that most of what is
downloaded is bulletins (think - large numbers of hams downloading the
same content as each other).
Unlike 20 years ago, broadband receivers (and sufficient computing power
for demodulation in software) are now affordable - a cheap RTL-2832
DVB-T USB dongle (plus preamp, filter, and antenna), can receive a 2.4
MHz wide chunk of spectrum on 6m, 2m, 70cm, or 23cm and your PC is
probably already fast enough to keep up.
So, I think that a usable ISS-PBBS would be cross-band full-duplex (up
on 2m, down on 70cm or 23cm etc), with asymmetrical bit rates - slow
uplink(s) which are compatible with ham's existing transmitters (1200
AFSK, 9600 K9NG/G3RUH, or maybe even n-QAM or OFDM via SSB?) and a fast
(broadband) downlink that can be received with cheap equipment (RTL-SDR,
Airspy, etc). Of course, data compression (LZW, LZMA, Brotli, etc) and
Forward Error Correction would be used.
Yes, a high speed broadband TX will require more power than a NBFM TX,
but presumably an ISS-PBBS would have a far more generous power budget
than a cube-sat.
"Spare" bandwidth on the broadband downlink would be used to constantly
stream PBBS bulletins (and maybe even personal messages when the
ISS-PBBS knows that the recipient is within radio range - see below re
grid squares), so that to get the latest messages, all you have to do is
listen (no ground station transmission required).
Thus, most users would only need to transmit when they have a bulletin
or personal message to upload.
I would have the ISS-PBBS co-ordinate channel access - ie ground
stations only transmit when the ISS-PBBS tells them that they can
(TDMA). The ISS-PBBS would have to reserve some of the uplink time for
connection requests, and there would of course be contention and
collisions between ground stations (which can not hear each other)
attempting to connect. To alleviate this, when the ISS-PBBS hears a
collision in the "connect request" time slot, it could perform a binary
search of the callsign space (as is done with some RFID systems - this
how they cope with multiple RFID tags within range of the reader at the
I think that it would make sense for a ground station's connect request
packet to include their grid square (Maidenhead locator), so that the
ISS-PBBS (which knows where it is on orbit) knows when that ground
station is out of radio range so that it can abandon attempting to
communicate with it. A simple enforcement mechanism (ground station
sends incorrect grid square) is that the ISS-PBBS would refuse to
communicate with ground stations declaring a grid square which is not
within what the ISS-PBBS considers to be radio range.
In areas of high amateur density, it will probably be more efficient
(and cheaper) to equip the local terrestrial PBBS with equipment to talk
to the ISS-PBBS, and then have the hams within the coverage area of the
terrestrial PBBS use that instead of the ISS-PBBS (this would reduce
contention for on the "connect request" timeslot of the ISS-PBBS, and it
will be more efficient for one PBBS to upload say ten messages to the
ISS-PBBS in one go than it would be for ten different hams to each
upload one messsage each to the ISS-PBBS).
This would give us global message forwarding that is not reliant on the