On 10 December 2014 at 13:09, Maria Blackmore <[email protected]> wrote:
>
> On 10 December 2014 at 02:51, Gord Slater <[email protected]> wrote:
>>
>>
>> In addition, two or more radio transmitters, or a single transmitter feeding 
>> several antennae, either by the most direct route or by adding delays or 
>> extra distance in the transmission line or signal path between the 
>> transmitter and receiver also cause interference patterns. They are utilised 
>> where needed and we try to avoid them where they would be a problem. We can 
>> even dynamically generate and manipulate these patterns in the time domain 
>> to provide performance gains or specific effects. That is far from a new 
>> concept. Look at a PAVE PAWS radar site for an easily-verifiable 1980's 
>> example.
>
>
> The several versions of the US "Space Fence" system also give an interesting 
> look at beam forming going back to the 60s. Look out for the S-band system 
> that they're going to be bringing online in the next few years.

Likewise the "Battle of the Beams" used a lot of different antennae
and receiver-path phasing methods for early conscious virtual
beamforming in WW2 as the research stepped up on both sides (rather
than just general directivity for the general sake of performance,
providing gain purely to reduce power considerations for example)
Sadly it isn't very mainstream stuff other than with general
descriptions - I've just had a quick google and Wikipedia only has 1
page on it with scant technical information on what was truly
groundbreaking research in several related subdisciplines.

Prototypes for what became the X-Gerat harness and radiators in
particular went through endless iterations as they tried to achieve a
suitably narrow beamwidth for the ranges required. I can recall some
of the ham radio or shortwave magazine articles in the early 90s going
into the systems in a little more detail as the Berlin Wall thawed the
Cold War slightly. Much of this topic has practical implementation
that dates back to the Y-stations in the early stages of WW1.

Mechanical attempts at near-real-time processing (with land-line voice
telephony as a backchannel to coordinate the multiple DF sites in
real-time for fixing) were tried, in similar manner to artillery
calculation aids, immediately after the phasing methods of additive
interference and especially, subtractive interference (due to the
deeper SNR nulls possible) became practical for 3-dimensional DF, thus
bringing the time domain into things. Now the phasing and cocked-hat
geometry is being done with general-purpose silicon instead of with
slide-rules and voice coordination on a backchannel and a naval chart.
Same game, just much faster, on transmit too, and much lower stakes
for those who were involved. This is, quite literally, 100-year old
practically-applied electromagnetics research 1914 and 1915 for the
inception of the techniques, with Frankie Adcock improving on the
then-dominant Bellini-Tosi phasing method of direction-finding
beamforming a couple of years later. Elimination (or at least the
reduction of) the descending sky wave component (again, by subtractive
phasing, leveraging the interference patterns in orthogonal/quadrature
separation and orientation) also improved accuracy around the clock as
the systems became more advanced.

Other methods of phase adjustment were attempted around the same time
but the Adcock performed the best. These were all quite conscious
attempts beamforming to provide optimal positional information to peak
or null the target station using mathematical methods of phase
adjustment, not just random observation of empirical cause and effect
or spacing, feeding, proximity and reflections - empirical methods
were also tried in parallel with various results)

We do understand the physics much more now and have the benefit or
some excellent calculation tools and display methods for design, but
it's identical in concept from a technical point of view - the aim was
still a 3-dimensional fix in as near-real-time as possible. In
particular, Zernike polynomials opened up faster visualisation methods
in the 50s that became very useful in radio and more general EM work
and not just optics.

Even disguising the base station appearance is far from new - in 1939
onwards it became quite important for some sites. Most of us nowadays
will have seen a tree disguised as a cell-site or vice versa:

https://www.google.co.uk/search?q=cell+site+tree
or bases and antennae attached to or behind extant street furniture,
for example. RF guys are forever trying to spot antennae and base
systems that are "nice" in one form or another, either covert or
overt. Many techniques rely on phasing to disguise the camouflaged
shape of the unit yet provide the required coverage pattern,
especially for downtilt of the beam toward the mobile targets and to
promote spectrum re-use in adjacent areas.
It's a geek's dream, "worse" than trainspotting and there are several
websites dedicated solely to it. There are clear social parallels with
Pylon Appreciation, another niche topic popular with some UK ISP staff
;)

> A "hifi" is not using interference patterns. In fact, one of the reasons why 
> there's a "sweet spot" in listening to a system with only two speakers, is 
> because you're trying to sit in a position which is a node at as much of the 
> spectral range as possible. You end up with parts of the room where you have 
> good high range response, mediocre middle, and good bass response, or 
> frequently were everything sounds good apart from bass frequencies.


Aye, not "using" them as such, but suffers from them as you describe
in detail later. I tend not to use audio analogies for wave
propagation but $generic_consumer_audio has the big advantage that
most people use it and can relate to it better than the RF
transceivers (intentional and unintentional - getting my usual EMC dig
in haha) that their "familiar" wave interference pattern concept
portable devices. Until recently it also had the advantage of having
two easily understood radiator positions (ignoring multi-driver
speakers for simplicity) and was similar enough in practice to help
visualise the electromagnetic . You can also bend the truth a little
and suggest the balance control is akin to adjusting the phase (and
not, in fact, the relative ERP of each speaker) to illustrate some
basic 2D steering concepts. 5.1 and bigger has made it all more blurry
though, on reflection :)

I'm very aware that though UKNOF/T is by no means non-technical, I
needed an example a bit closer to most people to tame down my rants
and concerns about practicalities and a real-world comparison that i
know many UKNOFers are very familiar with.

None of my criticism will be an issue in 5 years time when there's
enough of $everything to make the system work well at a very large
scale. At the moment, it's looking too much like a patentable concept
with salemsanship, when in fact it should be all that *with* provable
figures up front in a quick concise manner, that's all.

I'd love this sort of stuff to work well. But dressing it up as the
next big thing for the benefit of XYZ who can only handle executive
summaries (sorry - can''t find the right words) actually makes it
harder to believe, comprehend or trust at a practical level as a
hands-on RF guy. I used cocked-hat RDF (almost exclusively with SDR in
the last 2 decades and always with actively phased beamforming, *hands
on* almost every day of my working life in one form another until the
last few months. Reciprocity means it works on transmit too, though
different signal paths are needed above say 10dBW in my usage for
example.

 I still do use it occasionally, remotely (such is the power of the
internet these days) so I'm afraid see no real innovation in their
work. Maybe it's news to shareholders and investors and possibly users
if they can understand it, but from our point of view on the tech side
it's a passing phase that is rather annoying especially if it muddies
the extent of the achievements (ok, no more of the puns). After
watching the (almost) hour of video I had less clue than before I
started - the essence of my gripe.

With no tech FAQ or outline paper with measurements, their
presentations appears to me as the usual tiresome waffle from
marketers, which  (I am keen to stress about my rant) demeans the
actual work they've done and tech successes. In my mind even dismal
failure is a valid result of research of this kind - I realise that
not everyone would agree with that, especially (I'm guessing) their
target audience. There are many advances in electromagnetic research
yet to be made and we understand our inversal environments, consumer
mobile datacomms is only one outlet for the product of research,
"good" or "bad", positive or negative results are all worth it. If we
only concentrate on commercial success or let marketing and commercial
pressure control finance then the human race will lose out on so much
useful data.

Unfortunately, many people will read this and assume I'm talking about
aliens or free energy or morgellons :)

I'm not against SDR, centralisation of signal processing, or any of
the other "innovative" tech they are flag-waving about to their
audience, but having worked in the field (bad pun) myself for a long
time I've seen similar generalised claims over and over again but
*still* my 3G is almost unusable and my 2G voice drops as I move
around my town-centre home. Most end users have already shrugged and
given up - this should not have happened for anyone's sake. The extent
of my disappointment is hard to communicate - when I can't be bothered
to watch a video about phasing methods there's definitely something
wrong.
Yes, I rant, and to be fair my rant would have been better on the
UKNOT informal off-topic list (I initially thought it was, in fact,
sorry 'bout that!) I just want stuff to get better faster. I want the
tech to speak volumes, not the presentation get in the way. Usual
commercial pressure I suppose. It always is except in Original
Research.

If they had saved the fancy video clips for marketing and produced
reasonably full tech research for peer-review (even via Youtube, which
is better than nothing) I'd be selling it to people myself on it's own
merits. I know it will work, the physics is well known and as I've
alluded to, been used for a century in one form or another. I simply
think it's time we'd moved on from software methods of re-inventing
wheels and get on with inventing radical stuff, good or bad. The
"concept sell" birds-eye approach is simply so dumbed down it looks
like all the other snake oil and in my view has parallels with some
over-unity videos on YT and similar.

RF should not be a fringe specialism, with more research and respect
than it's currently getting. It's taken for granted by users and it's
particularly badly understood by the internetworking business
community as a whole, with a few exceptions, even though it is widely
relevant in that field in many unusual ways. It cannot be treated as a
wired method of delivery and the confusing collisions of terminology
are extremely unfortunate between the two disciplines.

We will all probably be carrying around  at least one device that uses
interference cancelling technology on both transmission and reception
of audio and that same device will probably contain up to 4 or 5
separate transceivers (or more likely one or more SDR ones instead)
and possibly two more radio receivers, both of whom use the
interference effect of phase cancellation and addition in
mathematically similar (though practically diverse) ways (FM stereo
and GPS. This stuff is everywhere.


-- 
sent via Gmail web interface, so please excuse my gross neglect of Netiquette

Reply via email to