[TowerTalk] Active phased arrays.

[Dudley Chapman]

Tom wrote.
 
That's exactly what the 1025 does.
 
I think what you are missing is the fact the signal can't be
separated from the noise and phased differently. The
combining of multiple phased antennas produces a pattern
that can be duplicated by any phasing system with 0-360
degree phase range and amplitude control. That pattern is
restricted by element spacing and phasing.
 
Tom,
   Yes, you are right about the 1025. And you are correct in saying that you
can't get any more patterns out of the array than you could if you used
passive components and painstakingly manually adjusted their values to every
conceivable useful value.  If you had a circular hex array of verticals you
could create a lobe in one direction and make notches in a number of other
directions by adjusting all the passive phase shift networks and the
complicated combining system.  In order to do that with mfj-1025s, you would
need six of them at least and it would take a lot of dial twiddling to make
them do what you want them to do.  
 
   Instead, if each of the hex elements had their own signal channel coming
in as a stream of complex numbers, you can combine them mathematically in
predetermined ways and even adaptive ways to create the patterns.  In fact,
you could combine them simultaneously in as many separate patterns as you
want in order to do something interesting with them such as noise
cancellation.  For example, suppose you have one pattern with a strong lobe
towards the signal of interest.  Simultaneously, you are computing another
pattern that has a strong lobe on the BPL power lines nearby with a null in
the direction of your signal of interest.  You use the second pattern to
filter and extract non-coherent noise information and adaptively use it in
an LMS algorithm to cancel BPL noise from the first pattern.  
   Separately, you also get the opportunity to do special demodulation and
signal detection using adaptive filters and other things that take advantage
of the coherence of the signal vs the random noise.  There are probably ways
to guide the heuristics of the adaptive filters by even using information
about the direction of the incoming desired signal vs signals that from
directions you are not interested in, and so forth.  And conversely, having
identified the signal, the beamforming can be more intelligently adaptive.
 
Jim,
   Yes, I completely agree with you on Moore's Law.  But it's become obvious
to me that a more general version of Moore's law (WA1X's law) applies to any
technology that is used for mass consumption of entertainment (which
includes personal communications).  Why is there more computing power on a
$200 graphics board than existed in the whole world 20 years ago?  It's
certainly not so we can do email and balance our checkbooks. Just consider
all technologies used to support networked video games, for example.
   Here is an interesting beamforming application just for keeping your
wireless LAN safe. It's a search and destroy system for network intruders
based on adaptive beamforming for direction finding.
 
http://hcac.hawaii.edu/tcwct03/papers/s25p03.pdf
 
WA1X's law is also adaptive. I am betting that as the population ages,
medical technology may also become a dominant driver for innovation.  It's
already true that MRIs have signal processing that we never dreamed of in
communications.  This field may not drive hardware prices down, but it will
certainly be the catalyst for some amazing new signal processing techniques.
Health, security, and fun.  Don't bet against these First World human
obsessions.  The economics behind these things guarantee that they will
produce cheap technology that can be used for all kinds of other
applications.
 
Dudley - WA1X