AT&T: Turning off Copper to More Than Half of Footprint

AT&T: Turning off Copper to More Than Half of Footprint
– 99% Coverage with LTE POPs
– Fiber to Businesses
By Dave Burstein of DSLPrime at

AT&T’s Project VIP, if approved by the FCC and state
regulatory agencies, means more than half of the US will lose
landline phones when and if Verizon follows suit. No cutback
like this has ever been proposed anywhere in the world since
Elisha Gray and Alexander Graham Bell invented the telephone.

About 1% of the population in AT&T’s footprint, about 20-30%
of the area, would not have wireless and would be in danger of
depending on satellite for every phone. Jim Cicconi, SVP at
AT&T, writes me, however, “No one we’re unable to reach with
broadband under this plan will lose their voice service.”

Some 20-25% will lose landlines and have to depend on mobile.
That’s probably half the land area, because the cuts will be
in the less dense areas. Fortunately, voice over LTE is almost
ready to take over and can have better call quality than the
old phone system.

It’s vital that the FCC ensure that prices are fair and
service is good. There’s an enormous potential savings if
wired networks are cut (shutting off the PSTN.) Major changes
are required to prevent negative consequences.

That won’t be easy. AT&T’s cheapest wireless plans are now
$70-$100, far higher than landline charges. The capacity of
today’s LTE means you can’t watch much video over them; it
could be many years before the greater capacity of LTE
Advanced and beyond are deployed unless required.

If consumers are protected, it’s right to go ahead.

Fiber to Half the Commercial Buildings in Territory
Businesses are willing to pay more than consumers, so
retaining them as customers is worth the $1-$3 billion
additional that AT&T committed. Cablecos have been gobbling
corporate customers for the last two years and offers 50-100
Mbps service cheaply. This was not a market for AT&T to give

The additional fiber will also make it easier to add wireless
small cells. Units well under $10,000 and soon closer to
$1,000 can supplement the coverage of the big towers. John
Donovan of AT&T has been talking small cells as the future for
years and this year they are beginning to deploy rapidly.
Landlords want the fiber to keep tenants happy. AT&T should be
able to convince many of them to allow small cells on their
roofs, perhaps without charging rent for the space.

No Real Increase in Consumer Broadband
First reports were mistaken. AT&T does not intend to “offer
its U-verse TV, Internet and Voice over IP plans in 8.5
million additional customer locations” for a total of 33.5
million. Don’t blame the reporters for that; the press release
suggested that interpretation. AT&T again and again has said
U-Verse had reached 30 million homes, including in financial
reports, where companies can be penalized for material
inaccuracies. In addition, AT&T has always planned to go to
~33.5 million in the near future. The current build contains
many irregular, irrational holes like much of San Francisco
and Indianapolis. They’d be stupid not to fill them in, and
John Stankey, chief strategy officer and group president of
AT&T, isn’t stupid.

AT&T found some way to characterize 6 million homes they’ve
been reporting as served and now unable to get service. One or
the other figure is wrong. It may be that the new figure of 24
million currently reached was achieved by subtracting 6
million current homes they need to bond a second line (already
in place) for best TV coverage. AT&T has said that they are
doing bonding since at least 2009 for those U-Verse homes, but
I haven’t seen orders for that many bonded modems hit the
supply chain. In either case, the PR is misleading. (Craig
Moffett and Todd Spangler also caught this last week.) Some
(possibly very low) percentage may get vectored lines capable
of “up to 75 Mbps” rather than the current 25-40 Mbps.

AT&T chairman and CEO Randall Stephenson also made much of the
plan to upgrade most of the 8-13 year old units to IP DSLAMs.
Those have long been obsolete and AT&T is already
systematically replacing them. Improvements in power, space
and manageability of the newer DSLAMs will likely pay for the
entire cost of the upgrade. Many customers will be able to get
downstream speeds of 5-15 Mbps instead of the previous limit
of six, which has been standard on most networks since 2003-
2006. A very limited number will be able to get much higher,
including some “up to 45 Mbps.” There will be many who can’t
get 5 Mbps and some two and below. Nothing was said about
upstream, but it’s probably very low.

LTE to 99% in Territory in Three Years
AT&T, knowing they would have to offer something to almost all
their present wireline customers, has decided to expand LTE to
all but the last 1%. The natural deployment, based on the
current towers and backhaul, would leave twice as many without
service. Verizon announced in 2009 that they would go to 98%
nationwide and AT&T was always going to come close. The
statements saying they would only cover 80% were for the
politicians. We all know about politicians and truth.

LTE Out of Territory to ~91/92%.
The 95% overall coverage leaves twice as many unserved as

The whole game in US wireless changed in 2009 when Verizon
announced that they would match their 2G footprint with LTE,
about 98% of homes. AT&T now announces that they would cover
the homes of 300 million people in 2014, when the national
population will be about 315 million. It’s 311 million now,
and increasing about 1% per year.

So their plan leaves 15 million uncovered, about 5%. If it’s
99% in territory (about 40% of the country), out of territory
would be about 91-92% covered. That leaves them somewhat
behind Verizon, but not too far.

The last 1% is 20-30% of the territory

AT&T’s man in DC, Jim Cicconi, writes “No one we’re unable to
reach with broadband under this plan will lose their voice
service.” That’s a crucial commitment, because otherwise a
quarter or so of the land area would have no voice service
except god-awful and expensive satellite. The fine points need
to be worked out, but that goes a long way to making this a
reasonable plan.

Chris Ziegler at The Verge was the only reporter I know who
realized the problem. “The biggest problem with that, it
seems, is that AT&T is only committing in its FCC filing to
offer broadband service to 99 percent of its current wireline
footprint, which leaves the other one percent — those in the
most rural, underserved areas of the country — in limbo, since
the carrier also wishes to decommission its legacy network.”

A good plan to protect that huge area is imperative.

Vectoring to 75 Mbps for a Few
From now on, every new build that’s mostly under 3,000 feet
should be vectored. Up to 300 meters, speeds are proving out
at 70-100 Mbps down, 10-40 up. AT&T hasn’t suggested they
would upgrade the existing 30 million lines, so only the last
10% of U-Verse is likely to benefit. It should add less than
$200 million to the overall U-Verse cost. Existing vendors
Alcatel-Lucent and ADTRAN hope for the contract, while Calix
is trying to edge in. Only Alcatel is publicly shipping
vectored DSLAMs, but ADTRAN and Calix are working to catch up.

In Germany, there are 12 million or more homes to upgrade;
Britain, France and Italy also have many non-upgraded lines
that would be natural. At this year’s Broadband World Forum,
it was clear that the technical problems are fast being
solved. But Britain and France are holding back on vectoring
because it’s not clear how it will work with unbundling. The
three countries have prices 30-50% lower than the US because
they have four wireline competitors. They don’t want to give
that up.

AT&T has for years had a contingency plan to use vectoring and
bonding if customers in volume leave for 100 Mbps from
cablecos. That isn’t in the data, and Randall publicly says
“20-30 megabits will be competitive for many years.” In fact,
AT&T has been holding market share in U-Verse areas where they
only offer 10-15 Mbps. To do a full upgrade to vectoring would
bring 70-100 Mbps to the majority of U-Verse homes for about
$2 billion a year for three years. That’s a plausible
investment, but AT&T has given no indication it will pull the
trigger unless cablecos get much more aggressive. U-Verse is
built to a 5,000-6,000 foot standard; vectoring improvements
fall off dramatically after 2,000 feet. So, many of the U-
Verse homes will not see speedups, even if AT&T moves ahead

Cablecos Are the Savior for High Speeds in More Than Half the
Abandoned Areas
About 92% of US homes can get cable modem service, nearly all
soon at 50-100 Mbps. Some 5-10% of the US has a broadband
problem, but unless you need more than 100 Mbps, you have a
highly capable connection.

If AT&T drops lines to 20-25%, as indicated, at least half of
those homes have a cable modem alternative. Many of them have
already done so; a key reason AT&T is giving up the territory.
That leaves 3-6 million homes in AT&T territory that will only
have LTE or satellite choices. These need special attention in

Wires are much faster than wireless for at least the next
decade in most places, and have far more capacity. So,
wireless is only a partial substitute. The price differences
with Europe make clear that the US situation of only telco vs
cableco is weak competition and leads to much higher prices.
That’s a market failure and even conservative economists
should be calling for government to do something.

The folks who influenced the final broadband plan —
Genachowski, Levin and Strickling — were strongly against
government action to compensate for weak competition. Unless
we get new, strong leadership, the US will remain among the
most expensive countries for broadband. The networks are good
except for the last few %, because we have cable to 96% and
Verizon jumped in 2009 with a 98% LTE build. But the prices
are inexcusable.

LTE Coverage Doesn’t Mean 5-10 Mbps for All
Verizon Wireless is consistently delivering 5-12 Mbps down in
most tests. But the usual maps of “coverage” are somewhat
exaggerated and include areas with weak coverage and hence
lower speed. Both AT&T and Verizon Wireless claim 100%
coverage of Manhattan, which I fact checked a while back with
AT&T. In reality, some dead spots remain and frequently areas
of weak coverage and sharply diminished speeds. (I believe
AT&T in particular has improved coverage from two years ago.)

The FCC wants to expand the objective testing of performance
to wireless. That’s even more important now, when many will
depend on LTE for broadband. AT&T strongly opposed the
testing, claiming the company reported all necessary
information accurately. Because I knew there were problems in
my hometown, I thought their claims likely unfounded. Now that
LTE performance is even more critical, I hoped AT&T would
abandon their opposition and the program move forward.

FCC head Jules Julius Genachowski measured actual broadband
speeds with the very effective Sam Knows measuring tool. OFCOM
co-developed it and produced a far more accurate picture of
Britain’s networks than ever before. The US results
contradicted what Genachowski and most of the pundits had
claimed. With one major exception, all the major DSL and cable
networks were reliable and generally delivered consistent
speeds. Decent DSL and cable networks are not congested and
have relatively few problems. Speeds don’t go far down even
during peak usage.

The exception, which almost no one I know expected, was
Cablevision. Apparently they weren’t doing a good job of
splitting nodes when needed. When the problem was exposed by
the FCC, Cablevision immediately corrected it. In the next
test period, the results were much better. This was a triumph
for Genachowski.

Sunlight is often the best disinfectant.

Voice over LTE is Better than Landlines; Data Has Capacity
Voice calls can sound better over LTE than they ever did on
copper, and far better than today’s mobile. The iPhone5 and
most quality mobiles today have HD Voice built-in, although
the US hasn’t turned HD on yet and may only experiment until
2014-15. Even HD VoLTE uses only a small portion of the
available capacity and can easily serve all the voice needs of
rural areas. Copper is more resilient in disasters and carries
power for the handset. Mobile failed miserably during
Hurricane Sandy and clearly needs upgrades for reliability.
AT&T is already selling a small box to connect your existing
“landline” phone to the mobile network; it’s a $10 add-on if
you have an AT&T mobile account and they are selling it
aggressively in Verizon territory. It might be worth including
a battery or at least a connection that lets users plug in
their own.

Data over LTE currently is severely restricted, with 2-10 GB
caps common. The speed is typically 5-10 Mbps down, a couple
up, and sometimes better. Since the average broadband user
already is drawing over 20 GBs per month, living with LTE
broadband is a major compromise. LTE Advanced, some of which
is already available, has 10x the capacity. Part of the
approval process for AT&T’s copper shutdown should be rapid
deployment of Advanced, realistic in 2016-2018. LTE Advanced
capacity can be 10x LTE’s capacity, although even that’s not
enough for heavy video use. Without a special push, most of
these areas probably would not get Advanced capacity until
next decade.

Sum up
If you take care of those not reached with LTE and price
capacity reasonably, it’s a good plan overall to eliminate the
copper. In theory, turning off copper is plausible. In
practice, it could very easily be disaster.

Bringing down AT&T prices is not an easy task for the current
FCC. All wireline competitors remaining will be cut out and
wireless competition for decent broadband will be minimal.
AT&T and Verizon are pricing LTE at twice the European norm,
according to the NY Times. A below average broadband user
drawing 20 GBs per month would pay $200 at AT&T’s current

If broadband for all is important public policy, those rates
need to be set by law or similar at a much lower rate before
this plan is approved. The FCC can’t be “technology neutral,”
but will have to insist AT&T be a pioneer upgrading to LTE
Advanced at a high profile and then rapidly move the unwired
areas beyond Advanced. The homes without wires will have a
seriously inadequate service in the current generation of LTE.

Fortunately, Bob Quinn of AT&T writes, “We know that there are
important policy issues that have to be addressed that are
just as important as the technical and technology questions,
and we are eager to engage in that discussion.” In fact, they
are deeply committed to this for business reasons so will do
what the FCC requires. Time for the commission to put
consumers first!

This article appeared in DSL Prime.

About the Author

The Online Reporter is the weekly subscription-based strategy bulletin about the enabling technologies of broadband, Wi-Fi, HDR, home networks, UHD 4K TV & OTT services; identifying trends in the Digital Media space. Only a fraction of our material here is published here. To see 4 free copies, follow the links above or go to

Be the first to comment on "AT&T: Turning off Copper to More Than Half of Footprint"

Leave a comment

Your email address will not be published.