[pskmail] Re: Speed adaptation in pskmail now complete
- From: Russell Blair <russell_blair86@xxxxxxxxx>
- To: pskmail@xxxxxxxxxxxxx
- Date: Sun, 3 Jan 2010 07:03:27 -0800 (PST)
Rein, Please set permission for me I cant get fldigi-3.13BL. Please dont for
get my puppy ISO with updates (fldigi-3.13BL server.0.9.30) Thanks.
Russell NC5O
1- Whoever said nothing is impossible never tried slamming a revolving door!
2- A government big enough to give you everything you want, is strong enough to
take everything you have.
- Thomas Jefferson
" IN GOD WE TRUST "
Russell Blair (NC5O)
Skype-Russell.Blair
Hell Field #300
DRCC #55
30m Dig-group #693
----- Original Message ----
From: Rein Couperus <rein@xxxxxxxxxxxx>
To: pskmail@xxxxxxxxxxxxx
Sent: Sun, January 3, 2010 5:51:54 AM
Subject: [pskmail] Speed adaptation in pskmail now complete
The adaptive speed adjustment feature for the pskmail arq system is now ready
for use.
The components you need for the java client are:
jpskmail-0.4.0.2
fldigi-3.13BL
The components for the server are:
pskmail_server-0.9.30
fldigi-3.13BL
**How does it work?**
The pskmail adaptive system automatically adjusts its packet length, speed and
mode depending on channel conditions. To allow this, the server constantly
monitors how successful a transfer is, and adjusts block length and modem speed
and mode to the fastest combination the channel will allow,
Three speeds/modes are available for each session, 10 mode profiles can be
chosen by the client at the the time a session is started.
Profiles available are:
9: PSK500-PSK500R-PSK125 (the fastest combination)
8: PSK500R-PSK250R-PSK125R
7: PSK500-MFSK32-PSK125R
6: PSK500-THOR22-PSK125R
5: PSK250-PSK250R-PSK125R
4: PSK250R-MFSK32-PSK125R
3: PSK500R-THOR22-PSK125R
2: PSK250R-PSK125R-PSK125R
1: MFSK32-MFSK16-PSK125R
0: Use server default
Fastest net speed attainable by the system is 41 characters/second (mode
profile 9), equaling almost 500 wpm.at a maximum bandwidth of 500 Hz.
This speed is reached transferring text on a clean channel without qrm/qrn and
a reasonable S/N ratio. Under these conditions each frame, consisting of 8
blocks transfers 512 payload bytes, with zlib compression. If binary data is
transferred the compression has little effect, and the throughput rate is 50%
of this value. (Pskmail is a text transfer system :)
Measurements were done with 1 Watt on 18105 kHz, the server is 20km away from
my QTH.
With this system the pskmail operator has a choice of modes for a variety of
channel conditions. A session will now easily survive a series of strong ALE
soundings, and even a longer pactor qrm session is no problem anymore. It will
just take a little longer....Block payload varies between 16 and 64 bytes.
The system uses RSID to switch modes. Actually RSID reception is the weak link
in the system, as the modes used are very robust, often more robust than RSID
decoding. Especially the new PSKxxxR modes with soft viterbi decoding and
interleaver are rock solid.
As the US based stations are not allowed to use PSK500, we had to take measures
against unwanted RSID switching. The new fldigi code (3.13BL) now includes a
table where one can choose which modes are allowed to be switched by RSID, and
likewise which modes are allowed to use RSID TX (tnx VK2ETA and W1HKJ).
The new server (0.9.30) has a provision which does not allow US calls to use
profiles 9, 7 and 6. If such a profile is requested the server will switch to
the default profile 5 (PSK250-PSK250R-PSK125R).
MARS users can overrule this limitation.
In EU we use PSK500R as default listening mode, but you can connect using any
mode available in the client.
The Intermar (maritime) servers listen in PSK250.
It is now also possible to use client-to-client connect for
keyboard-to-keyboard chatting with the jpskmail client. This chat feature does
NOT include the adaptive speed switching, as there is an operator on both sides,
Choosing PSK250R mode will give you PSK63 typing speed.
The system is full duplex and will transfer all typos without error....
Summing it all up, this opens quite some possibilities for experimentation,
which is what it is all about anyway, as far as I am concerned.
Have fun testing this...
73,
Rein PA0R
--
http://pa0r.blogspirit.com
Other related posts:
