[hfsig] RE: random byte stream compression - probably not

[Darryl Smith]

G'Day

GPS is not much use for data compression. Normal algorithms work well to
compress data so that it is 100% random. 

My thoughts on GPS is that I don't think it would help too much. From what I
know of HF the delays are quite variable [Hey, I am one of these people who
believes that a Worked All Bands award is for 50, 144, 440 and 1.2 MHz and
GHz)

If you look at http://www.johanforrer.net/SIMULR/ you will find that for a
simulated channel from Boulder, Co to Washington DC you can have a 1 Hz
frequency spread, and 2 mSec delay. But when the signal varies from good to
poor the delay varies a lot. 

If you are sending at a low bitrate I think GPS would help, but only where
each symbol was approx 10 mSec minimum. 

But then you have to ask yourself why you could not self-synchronize, which
is what you need to do anyway since GPS will give absolute time, not
relative time. 

Darryl

---------
Darryl Smith, VK2TDS   POBox 169 Ingleburn NSW 2565 Australia
Mobile Number 0412 929 634 [+61 4 12 929 634 International] 
www.radio-active.net.au\blog\ - www.radio-active.net.au\web\tracking
 

-----Original Message-----
From: Howard, Chris [mailto:HowardC at prpa.org] 
Sent: Tuesday, 28 September 2004 7:31 AM
To: Darryl Smith; hfsig at lists.tapr.org; Chris Howard
Subject: RE: random byte stream compression - probably not


It was mostly just me flapping my gums.

I was musing on some way we could use GPS
at both ends of an HF data conversation to
increase throughput, either as a time
sync or data compression tool.

What do you think of using the data flowing
from a GPS as kind of a sync stream
in parallel with the message data?
Would it improve error correction?

Chris


-----Original Message-----
From: Darryl Smith [mailto:Darryl at radio-active.net.au]
Sent: Monday, September 27, 2004 11:10 AM
To: hfsig at lists.tapr.org; 'Chris Howard'
Subject: RE: random byte stream compression - probably not


>Whether or not GPS/NTP time synchronization
>would help with improving data communication
>still remains.

Not sure what this is all relating to, but NTP/GPS can help decoding,
provided that the variations in channel delay are not significant compared
to the bitrate being decoded.

If the variations change too much (like on HF), then NTP will not help too
much. Coherent CW works well for moon bounce since the variant delay of the
channel is much lower than the size of the 'di' tone.

Darryl