The growth of municipal broadband networks

Mailman List Mirror (Read Only)
1

http://arstechnica.com/tech-policy/news/2011/03/133-us-cities-now-run-their-own-broadband-networks.ars

Ars Technica has a short article up about the growth of municipal networks, but principally a nice little 'hey check out this website' (http://www.muninetworks.org/communitymap)

The whole scenario around municipal broadband networks in a hopefully unbiased nutshell: Increasing numbers cities and counties seem to be getting frustrated with what they see as the lack of progress in broadband speeds from their incumbent provider(s) (even after incumbent provider(s) have been approached requesting faster speeds) and are deciding to do it themselves. Chattanooga, Tennessee has become the poster child for the idea, able to offer 1Gbps to users and businesses at competitive prices ($150 pcm.)

I'm curious how the feeling is on NANOG about shifting such provision towards municipal instead of corporations? I guess a rough summary of the competing views I've heard so far are:

+ It's fair and valid competition in the market, which is encouraging major ISPs to innovate instead of resting on their laurels and trying to do the bare minimum necessary to maintain their position and profits, an attitude that is stifling other economic growth?

- Local government is sticking its nose in where it shouldn't, providing unfair competition and stifling normal market processes. Municipalities are operating on the false belief that large bandwidth will automatically bring silicon valley to them, without understanding the bigger picture. That it's time, money and resources better spent on tax incentives or other means of encouraging businesses.

Paul

2

Paul,

http://arstechnica.com/tech-policy/news/2011/03/133-us-cities-now-run-their-own-broadband-networks.ars

Ars Technica has a short article up about the growth of municipal networks,
but principally a nice little 'hey check out this website'
(http://www.muninetworks.org/communitymap)

(snip)

I'm curious how the feeling is on NANOG about shifting such provision
towards municipal instead of corporations? I guess a rough summary of the
competing views I've heard so far are:

(snip)

With experience from Sweden, which has seen many varying incantations
of these sort of networks, I have this hopefully useful bit to share:
It's OK for tax-payer money to build layer-1 infrastructure if it
decides so, that non-tax payer money can sell services on, but fail
starts to happen the very moment they decide to go higher than that.

That's... all.

Regards,
Martin

3

Highly agree with this experience being shared. We have had some dealings
with municipal related fiber networks (not naming any names or giving any
hints for obvious reasons) where shortly after providing the proposal to the
customer, the municipal "sales vultures" went in and undercut our pricing
(with Internet access included) to below what the fiber loop itself was
priced at to us.

Paul

4

In a message written on Fri, Mar 25, 2011 at 08:31:21AM -1000, Paul Graydon wrote:

I'm curious how the feeling is on NANOG about shifting such provision
towards municipal instead of corporations? I guess a rough summary of
the competing views I've heard so far are:

If you look at the services going into most homes, what you find
are monopolies. Some are government run, some are regulated
monopolies, and there are now lots of hybrid models.

The funamental issue is that it is not cost effective to anyone
(governments, corporations, or citizens) to build multiple gas,
water, sewer, electrical, telephone or television distribution
systems to every home. Remember that when these services were
invisioned and first deployed telephone and television did not
compete.

It is only in very recent times that we have been able to overlay
Internet on both cable and television, and to have television
competition via satellite.

To that end, I think the US would be much better off with fiber to
the home on a single distribution infrastructure. That could be
owned and operated by the municipality (like the water system) or
owned and operated by a corporation granted an exclusive right to
service an area (think telephone, at least pre CLEC).

Where you immediately run into a snag is the next layer up. Should
the government provide IP services, if the fiber is government
owned? Should private companies be required to offer competitors
access to provide IP services if the fiber is privately owned?
There's a lot of space inside these questions for different models,
and I think there are at least a half dozen in play in different
communities.

Having looked around the world I personally believe most communities
would be best served if the government provided layer-1 distribution,
possibly with some layer 2 switching, but then allowed any commercial
entity to come in and offer layer 3 services. For simplicity of
argument I like people to envision the local government fiber agency
(like your water authority) dropping off a 1 port fiber 4 port
copper switch in your basement. On that device they can create a
layer 2 VLAN/VPN/Tunnel from any of the copper ports to any provider
in the town CO. You could buy video from one, voice from one, and
internet from another, on three different ports. You could buy
everything from one provider.

The actual deployments are a bit more complex, but I actually think
if in new construction we could drop telephone and coax in the
neighborhoods, and deploy fiber to the home it would be cheaper to
construct, cheaper to operate in the long term, and would end up
giving consumers a lot more choice. It is for all those reasons I
expect any established business to be firmly against it.

5

Oh, look. A hobby horse. My opinion is a third: that last-mile fiber is,
for several reasons, a Natural Monopoly, and having municipalities operate it,
and farm their residents out as customers to any comer to the meet-me room,
is the natural end-game ... and the sooner we get to it, the better.

The problem is that many states have made it *illegal* for muni's to do this,
encouraged, unsurprisingly, in many cases, by Verizon (who desperately needs
FiOS to fly, because they've cut-to-clear in their copper plant for so
many *decades* now that it's unsalvageable).

Check the archives; I think I brought this up myself about 6 months ago.

Cheers,
-- jra

6

+5

Cheers,
-- jra

7

And the natural question is - how will this differ from the way the
"government" services like water, power and transportation have
been run, privatised-but-not-quite, etc?

Adrian

8

Jay,

From: "Leo Bicknell" <bicknell@ufp.org>

Having looked around the world I personally believe most communities
would be best served if the government provided layer-1 distribution,
possibly with some layer 2 switching, but then allowed any commercial
entity to come in and offer layer 3 services.

+5

I've seen several cases of these types of networks rolling out the
MPLS cloud, oversubscribing ad infinitum, with lots of active network
equipment, which all in all in the end doesn't add *anything* more to
the end-user than hundredths or thousandths or even less of their
end-to-end link capacity, between them and the service-offering ISPs.

I'm very wary of doing more L2 than essentially required, and believe
it is much more sane to invest a bit extra in the L1, and skip
investments at this level in L2 entirely. Handing of L1 to providers
works perfectly fine, and adds no over-subscription. The only issue
with what I describe above is that it complicates the
multiple-vendors-over-the-same-pipe a little bit. Voice and video
works pretty fine over IP, though, last I checked. With a few new L1
network devices, the above should become even more feasible.
Convincing people they can build a network infrastructure without
switches is nearly fated for complete doom, though... (Perhaps giving
them some LED panels with high-power fans will satisfy their need for
blinkenlights?)

Regards,
Martin

9

It is only in very recent times that we have been able to overlay
Internet on both cable and television, and to have television
competition via satellite.

In "the old days" the phone company didn't provide "content". You
called someone and the people at each end provided the content or the
data going over the network. The phone company simply provided the
network. I still believe the biggest mistake we made was breaking up
the Bell System. We should have let them be, regulated the crap out of
them, and then said "no, you can't get into the business of providing
content". They system should have been left as a regulated public
utility.

To that end, I think the US would be much better off with fiber to the
home on a single distribution infrastructure. That could be owned and
operated by the municipality (like the water system) or owned and
operated by a corporation granted an exclusive right to service an

area

(think telephone, at least pre CLEC).

Yup, bring back "The Bell System".

Where you immediately run into a snag is the next layer up. Should

the

government provide IP services, if the fiber is government owned?
Should private companies be required to offer competitors access to
provide IP services if the fiber is privately owned?

I would say they provide network access only, not content. They would
be kept out of providing content and kept in the business of reliably
connecting content to consumer. That would be their focus.

Having looked around the world I personally believe most communities
would be best served if the government provided layer-1 distribution,
possibly with some layer 2 switching, but then allowed any commercial
entity to come in and offer layer 3 services.

I don't. What happens when the "government" then decides what content
is and is not allowed to go over their network? If one had a site that
provided a view that the government didn't like, would they cut it off?
I want the government very strictly limited in what they can and cannot
do and I want them to have to go to an outside entity for things like
lawful intercept because it is another check on their power. A private
entity might insist that there is a proper warrant or subpoena while the
government might simply decide to snoop first, get the paperwork later.
Keeping the network at arm's length from the government helps to make
sure there is another entity in the loop.

For simplicity of
argument I like people to envision the local government fiber agency
(like your water authority) dropping off a 1 port fiber 4 port copper
switch in your basement.

Big difference. Water is not a good analogy. The "content" in that
case is from a central source and everyone gets the same thing. With
the network, you have people communicating back and forth and much of
that communications is private or expected to be private (say, a phone
call or a secure financial transaction). If a private entity screws up,
it is much easier to fine them or fire the person responsible than it is
to punish a government department or fire a government worker. Besides,
we really don't need yet more people on the government payroll.

Though I do agree that it is a natural monopoly. It should be managed
by a regulated utility that is explicitly prohibited from providing the
content, only provide access through the network.

10

aka the "separation principle" ( Tim Wu - the Master Switch)

What surprised me is that when I put his point to Richard R.John at the
Columbia Big media event back in Nov
<http://isoc-ny.org/p2/?p=1563> - John totally agreed with it, citing the
precedent of the telegraph companies being locked out of the telephone
business back in the day.

j

11

I appreciate your argument.

When asked by Uncle Sam, the major RBOCs were apparently happy to hand
over customers' records and tap into their phones in direct violation
of the law. *Asked* not ordered by a court or any legally-empowered
person or entity. The companies and LEOs then had to fight for
RETROACTIVE PROTECTION FROM THEIR WILLFUL VIOLATIONS OF THE LAW, which
was granted by our federal legislature.

I think we would be far, far better off, from the perspective of
liberty, with a thousand small last-mile providers, some of which will
hopefully be owned by cities/counties/states and some of which would
hopefully be privately-operated. It's a lot harder to coerce (or just
ask) a thousand small access providers to block some "objectionable"
or "dangerous" content or activity without getting caught than it is
to do the same if there are only a handful of access providers.

Since there is no "liberty" advantage, in the real world, to a system
where AT&T controls the last-mile or states, counties, or private
contractors control same, I would choose the one most likely to create
a competitive business environment. We already know that homes
without cable television and Internet service are less valuable than
homes which have access to these services. I hope that communities
would develop and maintain the best last-mile networks they can in
order to attract businesses and residents with the most money to
spend, and the most to contribute to their tax bases, job market, and
skilled labor pool.

In an ideal world, I could agree with you. But you don't need a
tin-foil hat anymore to be absolutely certain that big brother has
over-stepped his bounds and will continue to do so even in an
environment where private businesses *could* be an obstacle. Guess
what, they aren't.

12

+1.

The layout of the old copper telephone layout is basically sound, you aggregate thousands of households via X length of cable into a single place, then you let anyone who wants to, rent space/power in there to put in their equipment and rent this fiber to the end user household/enterprise.

Do this with fiber to the home, and also provide rentable long haul fiber (to the next town etc) and rent out this infrastructure at decent pricepoint, you have a situation with a very low entry-cost for new players which is great for competition and in the long run, for the end customer.

13

The principle that kept telegraph and telephone apart wasn't a functional layering concept, it was a "technology silos" concept under which all communication networks were assumed to be indistinguishable from their one and only one application. If you read the Communications Act of 1934, you'll see this idea embodied in the titles of the act, each of which describes both a network and an application, as we understand the terms today. Wu wants to make law out of the OSI model, a very different enterprise than traditional telecom regulation.

14

I take your point, the separation was of a different order. But a
separation, nonetheless. The motive is not so much different.

I think we can all accept that "traditional telephone regulation" is rapidly
losing its grip as the beast morphs. Now that applications outnumber
networks new problems require new solutions.

I've heard Allied Fiber's Hunter Newby argue convincingly that really it's
about separating Level 0 - the real estate, the wires and the head end
premises - from everything else, and facilitating sufficient open access to
guarantee healthy competition in services.

And yes, where there's a monopoly there will have to some price regulation.
At least that's traditional.

As we've seen in the UK, while it's not so much a stretch to impose even
higher level unbundling on the telcos, when it comes to the cable industry
it's going to be a very painful pulling of teeth.
http://www.telecomtv.com/comspace_newsDetail.aspx?n=46077&id=e9381817-0593-417a-8639-c4c53e2a2a10

j

15

Whilst that's certainly true for some areas, it's definitely not the case
for all of the areas marked on that map.

The only entry for the SF Bay area is San Bruno, where the municipal-owned
cable provider *is* the incumbent, and has been for the past 30 years. Not
only are they the incumbent, but they are also a monopoly who have blocked
competition, resulting in higher prices than in much of the rest of the bay
area.

  Scott
  (Happily no longer living in San Bruno)

16

We aren't even suggesting that, George. We're suggesting that they *provide
access to people who provide network access*; we (most of us, anyway) don't
even think the muni's should provide IP routing. They should provide
*connectivity* to people who do that. And given how STBs work these days,
those wholesale customers could even be cablecos, in addition to telcos,
or IAPs.

Cheers,
-- jra

17

I think the motive for the traditional separation actually was
   completely different from the one for new separation. Silos had the
   effect of limiting competition for specific services, while the avowed
   goal of functional separation mandates is to increase competition.
   Opportunities for service competition between the telegraph and
   telephone networks were limited by technology in the first instance -
   you couldn't carry phone calls over the telegraph network anyway
   because it was a low bandwidth, steel wire system with telegraph office
   - to telegraph office topology - but you could carry telegrams over the
   phone network, but only if permitted by law.
   In a sense, ARPANET was telegraph network 2.0, and even used the same
   terminals initially. Paper tape-to-tape transfers became ftp, the
   telegram became email, and kids running paper messages around the
   office became routers switching packets.
   The layer 0 model has some merit, but has issues. In areas nobody wants
   to provide ISP services, and there is still a tendency toward market
   consolidation due to economies of scale in the service space.
   Facilities-based competition remains the most viable model in most
   places, as we're seeing in the UK where market structure resembles the
   US more than most want to admit: Their two biggest ISPs are BT and
   Virgin, the owners of the wire, and they have less fiber than we have
   in the US.
   Creating the conditions for network competition is a hard problem with
   no easy answers.
   RB

18

Again excellent points. And I agree, in the current UK model there appears
very little opportunity for independent ISPs to offer any significantly
improved service over the incumbent's own, and thereby grab market share.
It's all a matter of what else one can package with it - effectively the
separation principle anyway.

Creating the conditions for network competition is a hard problem with no

easy answers.

Where there's a will there's a way. The big question, to some extent is, is
there the will?

One doesn't miss one's water etc. I was cheered to see in the recent
Canadian usage pricing fracas, Marc Garneau handing out buttons saying "My
Internet Shouldn't Suck"[1], and also to see Susan Crawford urging students
to take to the streets over the issue [2] before it's too late. But it's
going to take the equivalent of 10 Tahrir Squares to overcome the incumbent
clout and establishment inertia.

Meanwhile we are seeing widening pre-emptive strikes like N. Carolina. the
incumbents ride roughshod over everyone stating words to the effect - if we
can't gouge we won't build..

There surely still have to be answers, however tough - and some kind of
separation would seem to be an inescapable component.

I am no techie, but alternatively I imagine what could be practically
discussed is how much new technologies like cheap plastic fiber driving
little wi-fi chips, mesh etc, could give those communities that haven't
already been legislated out of the game an opportunity to economically and
successfully build their own connectivity. [3]

j

[1]
http://www.theglobeandmail.com/globe-investor/crtc-wont-include-retail-services-in-internet-price-hearing/article1938694/
[2] http://isoc-ny.org/p2/?p=1930
[3] I am in the process of organizing a panel to discuss same at the INET NY
on Jun 14, expressions of interest welcome
offlist<joly@punkcast.com?subject=INET-NY>
.

19

It's all a matter of what else one can package with it - effectively the

separation principle anyway.

effectively negating the separation principle anyway.

20

+more

Owen