[PLUG-TALK] Government owned monopoly networks

[Keith Lofstrom]

On Fri, Jul 27, 2012 at 11:28:31AM -0700, Russell Senior wrote:
> While there is in theory no longer a monopoly (another
> company could build a second network), in practice the monopoly (or
> near monopoly) remains, and for structural/economic reasons, it is
> likely to remain so.

I don't believe the claim of "open access" on fiber.  Today, the
fiber from my house goes to a wavelength division multiplexer
somewhere nearby, where my photons join my neighbor's photons
on a fiber back to the switching office.  At the switching
office, the fiber plugs into a box that swallows all the photons 
and makes bits out of them.  There is probably a WDM somewhere 
inside the box, but there is no practical way to pull my photons
out of the box and send them to some other provider.   Ultimately,
I am talking to a single chunk of electronics belonging to one
organization.  Unlike copper pair, there is no way to punch my
fiber down to a different bundle to Aracnet or whoever, unless
they deploy a entire fiber network to every neighborhood WDM.

The physics doesn't change if we let some poor folks starve and
use the money to build a public fiber network instead.  There
is still only one backhaul to the switching office, and it still
goes to only one box there. 

There is an assumption here that the government will inherently
have more enlightened policies than a private company, and
spend more money to connect to competing providers with other
policies.  Enjoy your comforting fantasy - my reading of
history shows that companies can be good or bad, while
governments range from OK to mind-numbingly evil.

Any switching office, even if it is publically owned, can
still impose policies I can't abide.  For example, Multnomah
County has a library system with eBooks that are readable only
on proprietary Macs and Windows boxes.  Their library computers
are all Windows.  Their GIS is based on Silverlight, and only
Windows users can use that.  How many of their public documents
are Word or Powerpoint?  And this is looking from outside;
what do you Subjects of Sam have to deal with?

Washington County has its own problems (including disconnected
bicycle trails and bike-hostile drivers) but the IT is better
here.  Mostly.  I would rate both WashCo and MultCo as "OK",
a mix of good and mediocre and bad government behavior.

We moved to our current house to be close to Portland but within
Verizon (now Frontier) territory;  we are literally 600 feet west
of the service boundary with CenturyLink.

If you want fiber, with no restrictions on usage (so far), move.


> Fiber has the capacity to carry everyones traffic, for any/all
> ISPs/customers/insert-random-innovation-of-the-future.  It is
> idiotic to build more than one network in a particular area,
> as idiotic as  having seperate highways for different makes
> of cars.

So, you share your highways with airplanes, ships, railcars,
and transit rail?  It is precisely because interconnect
technology changes so rapidly that we MUST have more than one
network, hopefully with new technology or new design goals
for each one.  Single mode fiber indeed has high capacity,
but it can only plug into a single endpoint, and endpoint
infrastructure is evolving very rapidly. 

It may turn out that the easiest way to roll out IPV6 is to
provide a separate network with endpoints optimized for
IPV6 and ignoring IPV4.  Not because IPV6 can't share, but
because an IPV6-only network permits services and freedoms
that a mixed network cannot.

Imagine two or more networks with different security goals.  
Network A is the appliance network, locked down.  Windows
and Mac appliances, no programmability, walled garden content. 
Virus free and sterile - just the thing for the 99% of our
neighbors who are incapable of maintaining their machines. 
Network B is the inventor/programmer/new-technology network.
The inept are discouraged from participating.  It is
generative, expandable, complex, and free of DRM content. 
We get to try new stuff there, without worrying about
whether we are interfering with the delivery of American
Idol, or running afoul of the MPAA .

I give those as examples - things will probably split out 
quite differently.  Back in the early 1980s, Network A was
POTS and Network B (around Portland) was RainNet.  Then
Network B grew by a factor of 100,000 .  POTS is dying,
and we are fighting over what the Internet should look
like.  Forking the network into separate communities
with different goals and capabilities makes sense to me. 
We can still bridge packets where necessary, just as we
do with border routers now.


> Wireless will never *ever* compete with fiber in terms of capacity,
> principally because you don't have a good way of controlling and
> channeling the interference without the boundaries of a conductor.

I challenge that.  Wireless on a two dimensional surface with 
fixed omnidirectional antennas is indeed limited as you say. 
Wireless in three dimensional space with beamforming antennas
offers very high bandwidth, with no need for monopolies.  The
switches must be off the surface - in the stratosphere or in
orbit.  In the 38 GHz band (8mm wavelength), with 5 Gbps/channel,
a one square meter phased array antenna can see about 10,000
different patches of sky.   A 100 square meter antenna at
stratospheric height can see a million different patches of
ground, perhaps 100 square kilometers, or 100 meter square
patches.  So, we can deliver 100*10K*5Gbps per square kilometer.
Portland averages 2000 people per km^2, maybe 50 times that in
the high rise core.  So, worst case, we are dividing 5E15 bps
by 100K people, or 50 Gpbs each.  If that is not enough
bandwidth, we make our stratospheric antennas bigger.

Even though 3D beamforming radio can provide the bandwidth,
fiber is better for the high rise urban core.  And in that
core, it is cheap to deploy multiple fiber networks to each
high rise, especially if all those networks are doing different
stuff.  For neighborhoods with 1000 people or less per square
kilometer (most of the physical area of Portland), 500 Gbps
radio per person is more than enough, and a lot cheaper to
deploy and maintain.

How much radio power splatter is that?  About a microwatt
per square meter, sum total from all sources.  Omni antennas
are evil, beamforming is the future!

In a decade or two, radio will pass, replaced by agile laser
links.  We are one or two inventions away from cheap,
electronically steerable microradian lasers - we can already
phase lock large arrays of vertical cavity emitting lasers. 
So someday that phased array antenna becomes IR instead of
microwave, capable of deconvolving raindrops in the signal
path and talking through rainstorms without fade.

---

For a younger person, who hasn't seen very many technology 
revolutions, it is easy to assume that fiber + IP is the
be-all and end-all of communication systems.  Some of us old
farts remember teletypes.  Baudot teletypes.  Morse code.
110 baud modems.  Minitel-envy.  Arpanet.  Project Xanadu. 
And garages full of modems running PPP.  Some of my chip
designs were purchased by Cisco and others to route most
of the 1980s internet.  They are laughably obsolete now.

This stuff changes too fast.  A municipality with expensive
fixed infrastructure will protect that investment by locking
out new technologies that will make that investment obsolete
before it pays for itself.  Governments are like companies,
but with armed cops and jails.  A government will make the
very corporate monopoly deals that you fear - it has for
every other technology. 

You are looking forward to what you perceive as the
technology of the future, and assuming you can elect saints
to manage it.  I am projecting my mind forward, and looking
back at your obsolete technology and its police-deploying
managers with dread.

The government should stick to streets and sewers, water
and parks, schools and cops.  If they ever manage to get
those right, and have wads of spare cash and rooms full
of bureaucrats with too much spare time, perhaps we can
talk about adding to their responsibilities.  When that
happy day comes, I'll whip out my long list of proposed
new government responsibilities, with public networks
somewhere around #35 or so. 

There are people hurting out there.  The internet is 
mostly a hobby.  Please get your priorities straight.

Keith

-- 
Keith Lofstrom          keithl at keithl.com         Voice (503)-520-1993
KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon"
Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs