[Starlink] Intro and a question

Wed Jul 28 13:27:06 EDT 2021

  _____  

From: Mike Puchol [mailto:mike at starlink.sx] 
Sent: Wednesday, July 28, 2021 1:16 AM
To: 'Karl Auerbach'; starlink at lists.bufferbloat.net; dickroy at alum.mit.edu
Subject: Re: [Starlink] Intro and a question

I won't pound on the dead donkey too much, Dick is way above my pay grade on
this one, but I'll just add that multiple polarization schemes were explored
in the LTE world when they wanted to move from 2x2 MIMO (achieved with
cross-polarized antennas) to 4x4 and higher orders. They just couldn't make
it work, so efforts were abandoned - it's too hard to achieve the physical
separation the "invention" claims through any practical means, which is
where Carlos goes into ". and I'll need a big pocketed investor to prove
this works". Billions have been spent on research for CDMA, LTE and 5G, and
nothing in the polarization space has ever been marginally achieved.

[RR] The reason is once again the laws of physics!  Polarization is a
two-dimensional subspace of three-dimensional space . think Pointing Vector
= E x B, and you will rapidly realize claims of infinite capacity from an
infinite number of polarizations is nonsense.  As for not being able to make
4x4 work, it's because the people implementing it really do not understand
space-time signal processing applied to wireless telecommunications and real
antennas (and the same may be true of Starlink .. not enough available info
to tell yet:^((() And when you add to that confusing signal spaces with
physical 3-D space as Carlos does in his rant, you get BS on BS, or
BS-squared.  All that aside, what we really want, Dave, is the name of the
guy who invests in Carlos' scheme .  he/she obviously has money to burn!

Then, considering you only have a single path between a satellite and a
ground station, any claims of higher-order MIMO through physical separation
are void.

[RR] See above highlighted text! And just know that you are thinking that
physical separation has to mean "of the antennas at a single receiving or
transmitting unit" and there are "other physical separations that are
definitely relevant".  "Higher-order MMO" does NOT require large physical
separations however, just smart signal processing. Do it right and, under
the appropriate conditions/assumptions, the capacity of a channel of
bandwidth W between the satellite with M antennas and ground station with N
antennas) goes up by W*log2(M^2 * N^2) over that of a single antenna at each
end.   For example, if M = 16 and N = 4, the increase is 12W!  Point is:
MIMO does not necessarily mean multiple spatial streams, it means multiple
inputs and multiple outputs.  The two are NOT the same! He physical channel
between the inputs and outputs really matters.

RR

Best,

Mike

On Jul 28, 2021, 02:00 +0200, Dick Roy <dickroy at alum.mit.edu>, wrote:

Even better, this guy's BS goes back to the late 80's when Qualcomm was
promising 1000 times AMPS in cellular capacity with their new CDMA
technology! Fast forward, CDMA has disappeared from planet earth as a
cellular technology ... and 3 major telecommunications manufacturers are out
of business!!

RR

-----Original Message-----
From: Starlink [mailto:starlink-bounces at lists.bufferbloat.net] On Behalf Of
Karl Auerbach
Sent: Tuesday, July 27, 2021 4:09 PM
To: starlink at lists.bufferbloat.net
Subject: Re: [Starlink] Intro and a question

On 7/26/21 8:51 PM, Larry Press wrote:

https://spacenews.com/tech-breakthrough-morphs-gigabit-wifi-into-terabit-sat
ellite-internet/

<https://spacenews.com/tech-breakthrough-morphs-gigabit-wifi-into-terabit-sa
tellite-internet/>

Wow! I got massive deja vu and thought had to check whether I was
reading a 1970's copy of the JC Whitney Catalog!

(Or my own CaveBear Catalog - of hyperbolic bogus network stuff: "If we
have it, you don't need it"), most particularly:

- Our Press Release from one of the Interop shows, people actually
believed this!! Gaga Net:
https://www.cavebear.com/cb_catalog/techno/gaganet/ )

- The Maximum Momentum Router:
https://www.cavebear.com/cb_catalog/current/maxmoment/

(For those who don't remember, the JC Whitnet catalog was filled with
things for your car that would improve gas milage by a zillion percent
or bump horsepower by 200hp. With that catalog one could turn an old
1200cc VW bug into a flame breathing monster, or so one would if one
accepted the hyperbole.)

OK, let's accept this guy's claims as true. Do they make an end-to-end
difference?

Perhaps if the satellite part of the end-to-end path is truly a bit
synchronous "bent-pipe". But Starlink seems to be evolving far past
that simple bit-clocked-circuit model into something more resembling a
space internet with routers, or at least a switched network that could
have issues such as choice of route, multiple inputs feeding into one
output (in other words, potential congestion).

BTW, I did like the article's phrase "High Definition Internet" - It
immediately called to mind "Brawndo - it's got electrolytes" (from the
movie Idiocracy.

--karl--
_______________________________________________
Starlink mailing list
Starlink at lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/starlink

_______________________________________________
Starlink mailing list
Starlink at lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/starlink

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.bufferbloat.net/pipermail/starlink/attachments/20210728/566bbdf4/attachment.html>