[hfsig] Is there anyone out there?

[Charles Brain]

Hello Johan,

Well I managed to get unsubscribed somehow.
It is nice to have a place where we can talk technical.
Most of the other groups degrade into requests of how to
connect XX radio to YY interface to ZZ program.

By the way I think I finally managed to get the 
data directed equalisation to work (as per the Harris patent).
Although I suspect I have a bug somewhere in my code as it is 
not quite as good on weak signals as my RLS based equaliser.
On strong signals the constellation looks much better than the RLS
one. 

My RLS equaliser trains on the known data either side of the unknown data 
whereas the DDE uses the known and unknown data to update the channel 
impulse response and I think that is why it fails over on weak signals as it gets 
poisoned on the bad decisions.

What makes me think I have a mistake is that I went through the maths for the
Generalised Durbin Levinson routine I use to invert the autocorrelation matrix 
of the channel impulse response and I end up with an unexplained minus sign!

In the code if that minus sign is in there it doesn't work, if it is not it works!

- Charles
 
 
----- Original Message ----- 
From: "Johan Forrer" <jbforrer at yahoo.com>
To: "TAPR HF Modes SIG Mailing List" <hfsig at tapr.org>
Sent: Monday, October 12, 2009 8:15 PM
Subject: Re: [hfsig] Is there anyone out there?


> Hello Charles,
> 
> Am afraid, it has been rather quiet on this list.
> Guess everyone moved on to other projects?
> 
> Regards.
> 
> Johan.
> 
> 
> --- On Mon, 10/12/09, Charles Brain <chbrain at btinternet.com> wrote:
> 
>> From: Charles Brain <chbrain at btinternet.com>
>> Subject: [hfsig] Is there anyone out there?
>> To: hfsig at tapr.org
>> Date: Monday, October 12, 2009, 10:42 AM
>> Hello All,
>> 
>> I see there has not been much activity on hfsig for a
>> while.
>> This used to be 'the' place for interesting threads.
>> 
>> So is there anybody still out there?
>> 
>> - Charles G4GUO
>> 
>> 
>> _______________________________________________
>> hfsig mailing list
>> hfsig at tapr.org
>> https://www.tapr.org/cgi-bin/mailman/listinfo/hfsig
>> 
> 
> _______________________________________________
> hfsig mailing list
> hfsig at tapr.org
> https://www.tapr.org/cgi-bin/mailman/listinfo/hfsig