From: "Paul Keinanen" <keinanen@sci.fi>
To: rsgb_lf_group@blacksheep.org
Subject: Re: LF: Very large adaptive rx array
Date: Mon, 12 Feb 2001 09:18:04 +0200
Message-ID: <t41f8tcmjvitgn6kjnr8nad53mq6ue0bt4@4ax.com>
References: <94.ffdecc5.27b8a16f@aol.com>
In-reply-to: <94.ffdecc5.27b8a16f@aol.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

On Sun, 11 Feb 2001 21:16:15 EST, MarkusVester@aol.com wrote:

>Hi group,
>
>is there more you can do with that CW carrier-component than simply detect 
>the presence of a signal without being able to decode it? Well, we could use 
>it to guide a large adaptive array to regain the data rate:
>
>- 100 receivers in different locations across Europe record audio files of a 
>highly desired DX-station transmitting 3 s qrss, buried -10 dB in the noise,

It seems that you are thinking about similar array than the VLBI (Very
Long Baseline Interferometry) used in radio astronomy to get a high
angular resolution.
 

>- offline, we extract the frequency offset and phase of the DX carrier 
>components in each file using low bandwidth (e.g. 3 mHz). We also estimate 
>their signal-to-noise ratios;

In VLBI a very high stability frequency standard (hydrogen masers
etc.) is used as a phase reference on each receiving station and these
timing signal is inserted on the same tape as the received signal. All
tapes are read out by the correlator and the timing signal is matched
with those from other tapes.

A low cost amateur solution could be to record a commonly received
signal such as Loran-C, MSF, GPS etc. on the other channel of the
sound card, while the actual DX-LF signal is recorded on the other
channel.


>- using this information as a phase reference, we convert each file to 
>baseband (like in a synchronous AM-detector),
>
>- then add up the baseband data from all files, weighted by their SNR. Using 
>the carriers as pilots, we effectively have focused our large sparse array on 
>a small area around the transmitter, gaining some 20 dB in signal to noise 
>ratio,

In VLBI, the individual antennas have usually a beam width of a few
degree beamwidth. Thus the synthesised beam can be generated somewhere
within the individual antenna beams. Signal sources well outside the
beamwidth of individual antennas are attenuated, thus, reducing the
receiver overloading risk.

However, the individual antennas in a large LF array would be more or
less omnidirectional, so the risk for overload (or quantisation of
weak signals) is larger. If I understand the situation correctly, more
omnidirectional antennas would be required to synthesise a _single_
narrow beam than when using individual antennas with narrow
beamwidths.

With an array spread over 2000 km, would in theory produce a beam
width of a few arc minutes. However, since the propagation
characteristics are not stable on LF (compared to empty space), such
narrow beams would be more or less useless.

An electronically controlled array could also be useful for nulling
out some spot interference sources, but unfortunately most natural
noise sources are more or less omnidirectional, so a better strategy
is to put the maximum of the response towards the desired station.

>
>- and we finally extract the information at full data rate in its original 
>0.3 Hz bandwidth, saving a factor of 100 in time and transmitted energy.
>
>Just a thought after a glass of wine, or maybe good for something?

It might be a good idea to check what the radio astronomers have done
during the last decades and hopefully some good ideas could be found
that could now be implemented using low cost generally available
hardware and software (which is an absolute requirement for any
coordinated effort between several hams :-).

Paul OH3LWR