Starlink Satellites Observed

On 2020-04-26 12:32 PM, JF Mezei wrote:
On 2020-04-26 11:20, David Spain wrote:

uniform. It will have to make routing decisions based on load. YMMV when
it comes to latency.

The constantly changing routing decisions are what will really impair
the service, especially for stock trading as this introduces jitter
(constantly changing latency).

Stock trading lends to some seriously funky network topology. I've heard
some wild stories about networking companies routing traffic through
spools of fiber optic cable in order to lawfully keep local traders from
having unfair advantages over distant ones. There a technical term
traders use for this gaming of the system. I can't recall what it is.
(Front running the market?) It isn't a new concept.


Also, a "router" itself introduced latency not only because of the time
to process a packet (make routing decisions and send it to appropriate
interface) but also because it needs to fully receive a packet before
processing it, as opposed to a repeated that is an analogue device that
blindly repeats analogue pulses without trying to understand or process
them.

I've worked with this gear and written design verification tests for it.
You are being wildly pessimistic about the speed of this gear. Even the
power restricted gear that sits in a satellite. You are talking
minuscule decimal fractions of a millisecond worse case.


Constantly changing connections also introduce a lot of data overhead
because each router need so contanrtly update its routing tables to know
which satellite ad ground stataion is availab;e to its interfaces.

See above. The processors in this gear cycle in picoseconds.

Once you have satellite to satellite links, that satellite over London
will have route back down to the london ground station to pickup
traditional Internet to New York, and will have routes via multiple hops
across satellites to one which is above New York and then use that one
to connect to a new york ground station. There is logic in routing
algorithms to "prefer" certain routes/providers over others.

Yes, but the same is true for terrestrial connections. Not sure I follow
your argument here.

I'm not trying to be a SpaceX fanboy, but thanks to Jeff's link I think
I have a better understanding of how the Starlink topology works and for
sat-to-sat it appears to be reasonably competitive.

Dave

On 2020-04-26 1:57 PM, JF Mezei wrote:
On 2020-04-26 13:07, David Spain wrote:

Rural cell towers are well over 100ft in height, some of the ones I've
seen. If 5G is line of sight, that's a goodly line. But the common
freqs. probably will be the ones you say.

5G is NOT millimetre wave. 5G will be deployed on conventional
frequencioes first. The mmwave stuff is just lobby hype to make it look
essential for survival of ecomnomy for carrierts to be allowed to
install antennas for free and without eavy regulatory burden on urban
furniture.
IDNST. I'm providing an instance where mmwave could span longer
distances. Obviously 5G and mmave are not the same and I said as much by
agreeing that lower freqs. will be the more common ones. Why are you
putting words in my mouth?


I think you are being overly pessimistic. I WAS a user of GEO sat on the
Ku band so I don't have to speculate. Rain fade was extremely
occasional.

Heavy web snow will also kill it, and will accumulate and stick to
antennas.
You saw what I said about that based on real experience. Not a
significant issue. But go ahead and imagine otherwise.


It had to be raining pretty darned hard to kill it.

Average thunderstorm will kill it. Not for long, but will kill it.


Depends on where that thunderstorm is relative to the position of the
sat in the sky. Not every storm did for me. To say "average" doesn't
make sense to me. Also a factor to consider you don't mention, and that
is multiple sats are visible to Starlink pizza box at once. That wasn't
true of my GEO Ku band setup which was aimed at a spot target. How badly
do storms effect GPS? Might be a better comparison than to GEO sats.

BTW my GEO set up also suffered from solar glare twice a year for about
20 minutes each time as the sun passed behind the sat and obscured its
signal. That won't be a problem with Starlink.

Dave

In article ,
says...

On 2020-04-25 18:39, Jeff Findley wrote:

In the interim, the beta testers/customers that are within range of a
Starlink satellite to ground station hop will "just" have satellite
Internet that has lower latency than regular GEO satellite Internet.

Everything will beat GEO Internet service.

And there is your primary business case right there.

But if Starlink wants to
proide Starlink service in Canada's acrtic, it will need to build a
ground station there, and that ground station will need to use GEO
satellites to link the ground station to the Internet.

Hell of a lot less customers in Canada's Arctic than there are in the
continental US. Starlink will no doubt focus on the most profitable
markets first. Some of those markets simply do not need crosslinks.

The one thing LEO constellations share with LEO is the uplink capacity
limits which is limited by the spectrum they are allowed to use betwen
satellites and ground stations.

We shall see. But do note that since the satellites are much lower than
GEO, it takes a lot less power for a ground station to transmit to them.
The equations are not as simple as one would hope.

https://www.electronicdesign.com/tec...ns/article/217
96484/understanding-wireless-range-calculations

The more ground stations you have (with
focused beams for GEO or just limited line of sight for LEO) the more
capacity you have. SpaceX may be able to launch thousands pizza boxes
for free, but it still has to deploy a high number of gound stations,
each providing adequante connectivity to the Internet.

Still orders of magnitude less ground stations than the number of 5G
towers needed to widely deploy 5G. This is why 5G won't be serving very
many rural customers (i.e. potential Starlink customers).

But if you wish to serve a farmer in Otumwah IOWA, you may have your
Chicago ground station that is in range. But Otumwah, if it doesn't
already have FTTH or even DSL, is likely to have cellular. So much of
the footprint aorund a ground station may already have service.

Not 5G cellular in the sticks. Again, I know people that live "out
there" and they're lucky if they have DSL which isn't sufficient for me
working from home like I am now.

And once you have inter-satellite links, the more satellites are
aggregated into one that as sight of a ground station, the less capacity
you have per user.

This is bull****.

LEO solves latency, not capacity.

This is also bull**** when you compare to GEO service. Look at the
equation of RF power over distance. Look at the exponent on distance.

Jeff

--
All opinions posted by me on Usenet News are mine, and mine alone.
These posts do not reflect the opinions of my family, friends,
employer, or any organization that I am a member of.

In article , says...
There were also
at that time daily download caps there were severe. I can't recall now
exactly what it was but it was laughable by today's standards. Something
on the order of 300MB or so.


This is why my remote co-worker had to go to the office to do downloads
of the once every other week "full" builds. Those, at the time, were
typically 100 GB. These days, it's more than double that. The daily
incremental builds are smaller, but still measured in tens of GB.

Jeff

--
All opinions posted by me on Usenet News are mine, and mine alone.
These posts do not reflect the opinions of my family, friends,
employer, or any organization that I am a member of.

In article ,
says...

On 2020-04-28 08:01, Jeff Findley wrote:

Still orders of magnitude less ground stations than the number of 5G
towers needed to widely deploy 5G. This is why 5G won't be serving very
many rural customers (i.e. potential Starlink customers).

Once again, 5G is a protocol which can be deployed on any frequency and
will initially first go on the same frequencies as LTE, and likely going
to 600 for rural areas. mmwave is just to add downtown capacity but
won't be used outside of densely populated areas. You need to ignorte
all the lobbying hype that has been done , in particular to convince
FCC/FTC to allow the T-Mobile attempt at saving Sprint.

The problem is not 5G technology as much as the fact that the mobile
companies don't profit as much from upgrades to infrastructure in rural
areas (except along interstate highways). Because of this, there is
less incentive to upgrade rural infrastructure. So, rural areas often
lag behind cities when it comes to basics like signal strength and
bandwidth (which are intertwined since you can't have high bandwidth
without decent signal strength).

My mom lives out in the country. She's stuck with crappy DSL and very
marginal cell phone coverage for my carrier. It's like being stuck in a
data desert when I visit. My smart phone is pretty much useless there.

Jeff

--
All opinions posted by me on Usenet News are mine, and mine alone.
These posts do not reflect the opinions of my family, friends,
employer, or any organization that I am a member of.

In article ,
says...

On 2020-04-30 11:32, Jeff Findley wrote:

The problem is not 5G technology as much as the fact that the mobile
companies don't profit as much from upgrades to infrastructure in rural
areas

Upgrading an existng radio to the next generation does not cost that
much, and it allows the carriers to streamline site management to the
same software. The new generation (5G in this case) has better
compression so gives more capacity for same amount of spectrum. So it is
to the advantage of the carrier to upgrade.

This is not the same as adding more cell sites. I am talking about
upgrading existing ones.

Then why is my cell reception absolute crap in the country if it's so
cheap to implement high power high bandwidth towers in rural areas?

Answer: $$$$

Jeff
--
All opinions posted by me on Usenet News are mine, and mine alone.
These posts do not reflect the opinions of my family, friends,
employer, or any organization that I am a member of.

On 2020-04-30 11:32 AM, Jeff Findley wrote:
My mom lives out in the country. She's stuck with crappy DSL and very
marginal cell phone coverage for my carrier.It's like being stuck in a
data desert when I visit. My smart phone is pretty much uselessthere.

Jeff

On 2020-05-01 9:32 AM, Jeff Findley wrote:In article
,
says... This is not the same as adding more cell sites. I am talking
about upgrading existing ones. Then why is my cell reception
absolute crap in the country if it's so cheap to implement high power
high bandwidth towers in rural areas? Answer: $$$$ Jeff
Yes, but there is one other consideration. Not all cell carriers are
equal in where they choose to provide service. My old hometown is not
well served by the carrier I use for my smartphone at home, in a
completely different state. But were I to switch carriers, I would have
seen a vast improvement. Jeff, the same might be true for where your mom
lives. Or not. Just saying it is a consideration. YMMV.

Dave

Let's try that again:

On 2020-04-30 11:32 AM, Jeff Findley wrote:
My mom lives out in the country.Â* She's stuck with crappy DSL and very
marginal cell phone coverage for my carrier.It's like being stuck in a
data desert when I visit.Â* My smart phone is pretty much uselessthere.

Jeff

In article ,
says...

On 2020-04-30 11:32, Jeff Findley wrote:

The problem is not 5G technology as much as the fact that the mobile
companies don't profit as much from upgrades to infrastructure in rural
areas

Upgrading an existng radio to the next generation does not cost that
much, and it allows the carriers to streamline site management to the
same software. The new generation (5G in this case) has better
compression so gives more capacity for same amount of spectrum. So it is
to the advantage of the carrier to upgrade.

This is not the same as adding more cell sites. I am talking about
upgrading existing ones.

Then why is my cell reception absolute crap in the country if it's so
cheap to implement high power high bandwidth towers in rural areas?

Answer: $$$$

Jeff


Yes, but there is one other consideration. Not all cell carriers are
equal in where they choose to provide service. My old hometown is not
well served by the carrier I use for my smartphone at home, in a
completely different state. But were I to switch carriers, I would have
seen a vast improvement. Jeff, the same might be true for where your mom
lives. Or not. Just saying it is a consideration. YMMV.

Dave