Hi Ben

Okay well there are a few basic possibilities here;

- Packet or AX25
More use in a multi station environment with interactive keyboarding
going on. It was probably the most serious attempt at a data mode that
happened to amateur radio - back in the late 1980's. (excluding CW ,
RTTY etc) Not great for file/image xfers as TX/RX times for fixed TX
lengths and frame retransmit error correction means lots of time is
spent in T/R switching and resending entire frames. The "original"
packet standards were 1200 and 300bits per second or about 85 characters
per second at the high end. Can be optimised to get around 110cps if you
play with packet lengths etc. It used FSK which by todays standards is
not a very efficient use of bandwidth. There is a LOT of software
available for it. Early work tended to use external hardware (A TNC or
modem hanging off a PC port) Nowadays there is a lot of s/w that relies
on the PC soundcard to do the DSP work in encoding/decoding whatever
mode you want. My perosnal favorite is Tom Sailors "soundmodem" code
that creates the "modem" in software and then talks to (or supplies) the
stack above that to talk vanilla AXC25 connect mode or TCP/IP etc. This
code is available for both Linux and Windows platforms. Lookup "Flexnet"
or "soundmodem". Good start point might be http://www.baycom.org. There
are other software packages (eg MixW?) that use the soundcard but their
names escape me at the moment.

- Narrow digital modes
These tend to occupy less than (say) 1khz b/w and work well with low
power/low s/n situations. Pretty slow for file xfers many indeed not
even allowing it without lots of user intervention. You may want to
lookup PSK31, MT63, Olivia and so on. I also note someone has modified
some PSK63 code and i/f to talk email through the normal clients. PSK31
for example is only 31 bits per second! I would not recommend any of
these modes for file xfers unless you were fighting low signal levels. I
for example am going to try some low rate beaconing of GPS data some
day, hoping to get reliable info over maybe 100-200 miles on 2m SSB.

- RDFT Wyman encoding
This was originally touted as a "perfect" SSTV mode but in reality it
sends any form of digital data. The code can be run both under Linux and
Windows as a command line utility. There are a few front ends that make
use of the underlying code (eg Digtrx). The code itself comes in two
parts. One encodes the binary data to a wav file that can then be played
with any media player through the PC speaker port to the radio input.
The other takes a wav file as an input and converts it back to the
binary data. The front end programs tend to make this part easy as it
recognizes the start of the stream to start recording said wav file. In
its original form the RDFT code was made to work over SSB bandwidths
with varying degrees of forward error correction (ie it doesnt need to
retransmit because of errors) Data rate is in the region of about 85cps
which is real good compared to SSB on the original FSK packet. I have
personally played with the code and managed to get a 50% higher data
rate over a FM voice channel. There was a rumour that the author was
looking to make a faster executable for that purpose but I havent heard
anything for a while. The original code description (at
http://www.kiva.net/~djones/digmodes.htm - only Wyman 11-13 have been
released) talks about character rates in the order of 250/sec. There are
of course some limitations to this code. Maximum binary data size that
can be sent in one session is 64Kbytes. No drama really as you could
break files up before sending. It also requires a lot of CPU and memory
grunt to decode. We are talking humungous memory use here, so much that
it is often worthwhile decreasing binary filesize to less than (say) 20K
chunks so that the OS doesnt swap RAM to disk and slow the whole
process. From memory a full 64K file uses in the order of 800MB-1GB of
RAM. What I am trying to get at is that it would be real difficult on
older PC equipment or a PDA.

I should point out by the way that RDFT is also available as a modem
inside of packet and can supposedly do 2500bits/sec on HF/SSB. This is
available in the soundmodem/flexnet package. It is called Q15X25 or NEWPSK.


It would take me a lot of time to assemble all references to software so
I am hoping the above will steer you into a specific direction. If you'd
like to expand on your possible usage I might be able to steer you
better. I'd need to know what kind of radios you are talking about (eg
FM vs AM, VHF, HF, intended use/range).

Cheers Bob VK2YQA

wrote:

Bob:

I agree, its not worth modifying the walkie talkie if a
data-to-voice encoder and voice-to-data decoder exist
that can reasonably transmit over a walkie talkie's
channel. Would love to review some of references on
software or studies done already. Thanks so much in
advance.

Ben