Copyright © 2001, CIMI Corporation, Voorhees, NJ, USA. All rights reserved. This issue may be printed or stored in Web format for personal, non-commercial, use, provided that the entire issue including this copyright notice is reproduced and included. Portions of this issue may be quoted, printed, or stored providing that the subject portion is annotated with the issue identification above, and is attributed to the copyrighted material of CIMI Corporation. Other publication, reproduction, electronic storage or retrieval of this material, in whole or in part, without the express written consent of CIMI Corporation, is prohibited.
This section is always reserved for subscribers. This month, it covered service provider CapEx growth and trends.
First, we want to
take a moment to codify something that’s been empirically true for a
couple of months; the Management Briefing segment of this
newsletter will be officially dedicated to recounting the important
events of the month. Some
months, that may mean focusing on a single happening, and in others it
may generate a laundry list. We
hope this will be useful to our busy readers, many of whom have asked
for a general status review.
During the last
month, we’ve continued to make progress toward telecom reform.
The Tauzin bill was read and amended in both the Subcommittee on
Telecommunications and the Internet and in the Energy and Commerce
committees. The next step
should have been a vote in the House, but the bill was diverted to the
Judiciary committee, apparently to reconcile it with other legislation
there aimed at linking RBOC failure to comply with the Telecom act to de
facto anti-trust violations. It’s
hard to see how either of the Judiciary bills would be incorporated into
the new legislation, and even if they were it’s not clear that the
effect would be negative.
The amendments
offered to the Tauzin bill itself were interesting and not particularly
debilitating to its intent. The
most significant were:
1.
A change to require a
specific level of DSL empowerment over a specified period, with progress
points starting after a year. This
amendment brings the bill into better correspondence with the Senate
bill of 2000, and its introduction might indicate some deal-making
between the House and Senate to facilitate passage.
2.
A change to explicitly bar
the application of the FCC line-sharing order to fiber remote
provisioning. This seems
directed at eliminating a perceived conflict between the packet
exemption order (99-238) and the line-sharing order (99-255), with the
former exempting packet infrastructure from unbundling and the latter
requiring RBOCs to sell only the high-frequency spectrum on the loop.
99-255 would still apply in the context where it’s nearly
always applied—to home-run copper.
3.
A change to exempt current
interconnect agreements from the impact of the legislation.
This is consistent with previous FCC actions, which have always
contended that interconnect agreements are contracts and are not
automatically invalidated because one or more terms which had been
forced by statute are no longer forced.
4.
Some wholesaling of
services of the “high-speed data network” is required, but it
appears the pricing basis for that wholesaling will be more rational and
that the term will be limited to three years.
This amendment appears to reduce what was effectively a
“windfall” for the RBOCs, who under the previous text were immune
from even offering wholesale rates on the new services.
Still, it may be the most troubling in terms of impact on RBOC
ROI.
We do have some
concerns arising out of the current process, albeit minor ones.
First, we are
concerned about the explicit linkage of data service reform and access
reform to the “Internet”. Just
what is that? Does the
Internet mean routed IP service? Does
it mean services from current ISPs?
Does it mean any service that is given the name?
There is no really good legal definition of what the Internet is,
and the definitions provided in the bill itself tend to lean in the IP
direction. We’re afraid
that this might be interpreted as excluding some forms of video or
real-time content delivery that didn’t happen to use IP packets.
In any case, it seems an invitation for further clarifications in
FCC and court rulings, and this could defer RBOC willingness to act
aggressively on the measures finally passed.
Second, we’re
concerned that the concept of unfettering the RBOCs for data services
may be a bit contaminated by the requirement to wholesale those services
in any form. Again, the
problem is the way this will be interpreted and the way the RBOCs might
look at its impact on their ROI. There
is a delicate balance here, and it may be that the bill’s failure to
separate “data services” and “access network” cleanly is at the
core of the problem. Data
services infrastructure should be absolutely exempt from all wholesaling
and unbundling, as should anything the RBOC does out of region.
Access services should be available for wholesale access at least
for a time, but in-region only.
According to a
friend in the DC area, the politicking on this bill is reaching the
level of an election blitz, with people taking out commercials to
espouse their view. It seems
to be shaping up as AT&T versus the RBOCs, which also gives us
concern.
AT&T has always
been behind the drive to cut off RBOCs from new markets.
That’s understandable; everyone wants to win, and every
corporation has a responsibility to their stockholders to get the most
bang and buck for them. The
problem here is that the media seems to be taking sides as though the
issue is “fair competition” for the RBOCs, and the issue is really
one of whether the RBOCs will be forced to subsidize competitors who
clearly cannot make their business model work in a free market.
AT&T wants that, obviously, but it’s bad for the industry
as a whole.
AT&T is also
stepping up its battle at the state level, which seems to make the
Tauzin bill’s suspension of other regulatory jurisdiction on the
matter an issue of critical importance.
While AT&T lost in
SBC’s keynote
address at N+I was interesting in this light.
In it, the spokesperson talked about PON, fiber to the home,
trials of new residential fiber, and other “pie in the sky” kinds of
issues. This comes at a time
when many (including us) have reported that SBC’s Pronto architecture
isn’t content or even real-time capable.
That’s been true all along, of course.
So has the fact that SBC has intended to offer PON-based access
to both business and residential users.
Thus, the “new” disclosures are “new” not for what they
commit to (which isn’t new at all) but for the fact that they seem to
draw the eye of the market into a 2002 timeframe, away from the near
term. That reflects, in our
view, the fact that the regulatory situation is holding the near term
market hostage. SBC is
sending a signal to regulators—pass Tauzin or wait years for new-age
access.
The Senate, of
course, is now in control of the Democrats, who have apparently sided
with the CLECs and AT&T. Though
there seems to be a real effort at bipartisanship (the Tax Cut bill was
passed), it’s not clear how things that are politically low-key like
Telecom Reform will do. As
of now, we have to say that the best-case scenario of an August
recognition that telecom reform is inevitable now seems unlikely.
Then there’s the
stock market. The NASDAQ
went up with the Fed’s off-cycle rate cut, then gradually sold off as
reality set in. The May 15th
cut was largely factored into the market pricing, so the analysts say.
They’re apparently right, because stocks went pretty much
nowhere on the 15th.
From this, we’d hope the industry could learn a lesson; unless
the Fed does a surprise cut, nothing much happens, and there aren’t
enough percents left to do much more surprise cutting, even if Greenspan
were so inclined.
Here’s the truth;
rates can’t go more than one percent lower, period.
Even to get that low, we’d face a reactive rise of a half-point
or more within 6 months. Thus,
the Fed can’t stimulate the stock market, it can only stimulate the
economy. What’s worse (for
us in networking, at least) is that networking is only a part of the
stock market, and not the part most responsive to monetary policy.
If the Greenspan Genie gets things rolling in earnest, the most
likely sectors of technology to benefit are those at the consumer
level—PCs, chips, etc.
What about us
lonely networking people, then? The
problem we have with stocks is a balance of hope and fear.
Wall Street hopes networking will come back.
Wall Street fears that it won’t—that any “new” ideas for
networking success are simply rehashes of the same tired lies that were
discredited in 2000 and earlier this year.
The current Fed
cuts, combined with the tax cut that’s forthcoming, will surely
stimulate the economy overall. If
that stimulation isn’t accompanied by a collateral development of a
sane network profit model, the capital generated will go elsewhere and
networking will lie fallow (so to speak) for a couple of years until it
gets its house in order.
All this weighed on
N+I, and one report of the show talked about how vendors were getting
“more realistic”, which means (in reporter-speak) “abandoning the
utterly impossible in favor of the highly unlikely”.
There is still an industry view that the problems with the
paradigms of old are simply cosmetic, and can be cured with just the
right spin or the addition of a relatively minor ingredient.
Get real; this problem is going to require substantive
alterations in our basic mindset in order to develop a credible and
effective solution.
What will the
solution be? We’d love to
tell you, but the problem is that whatever it is, it will probably be
the child of reality and opportunity combined, meaning that spin will
certainly play a role. We
can define only the “reality” part; the rest is up to the market.
Sill, there are some contender approaches:
1.
Content networking.
Obviously this is the long-term revenue engine, and one we’ve
covered in a previous Market Area Focus.
The problem, as we said in our earlier review, is that content
networking is a relatively slow starter, with most of its benefit
hitting in the 2004-and-beyond period.
It also demands more of the access network, and thus is more
dependent on the exact nature of regulatory reform.
2.
IP VPNs.
If we define VPNs the way CIMI Corporation has always defined
them (which, defiantly, we’ll call “the right way”), the VPN
opportunity in 2001 is about $10 billion, of which only about 40% will
be realized. The remaining
money would certainly help at least some of the second-tier players—if
they could get it.
3.
Advanced voice services.
Things like unified messaging, custom calling, and tuned voice
service features are all revenue opportunities.
Custom calling represents a total opportunity in the $25 billion
range (worldwide), and sequential growth next year could be on the order
of 50%.
All of these things
demand that we start looking at public data services (or public
non-voice services) in a less Internet-specific way, perhaps even less
IP-specific. That may prove
to be a tall order for the industry.
We think that
market conditions will eventually provide the answer to “What next?”
The best-case August recovery is unlikely now, and the best
available solution in the regulatory forum would now take at least
another nine months to be felt by the market at all.
It’s not impossible that we’ll see another networking
collapse, driven by the optical sector but wide enough to bit most of
the players. It may be that
this crunch will be enough to spur the FCC or Congress into action.
By the end of the third quarter, it should be possible to see a
glimmer of how the future will unfold, but in the meantime we’ll help
you track the driving events.
Ok, we know our
readers are aware that we have a certain…well…cynicism about most of
the so-called “technology revolutions”.
For a number of reasons that most won’t find all that
important, the hype pressure has been strongest in the access space.
Revolutionary access, addressing as it does a recognized service
bottleneck, has a strong emotional appeal, and it’s a waste of time to
generate hype where there’s no visceral interest.
Broadband wireless
is squarely in that revolutionary access space.
To many, it’s the straw that’s going to break the back of the
huge, unresponsive, un-everything RBOCs.
Well, perhaps it is and perhaps it isn’t.
There are certainly situations where broadband wireless
technology makes a lot of sense. There
are also areas where it clearly will never recover cost.
The question we must ask is whether there are enough of the
former to build a business.
Much of this
material will be edited before being published on our website; the
market information and our views are for subscribers only.
Sorry, Internet readers!
Like most things in
our marketplace, the broadband wireless space is very vague in terms of
scope or definition. Everyone
agrees on the “wireless” part, meaning that the technology has to
work without any physical media between customer and network.
In short, it’s a free-space, radio-frequency, transmission
system. Even the term
“broadband” is sort of agreed upon.
We’ve come to accept the FCC definition that “advanced” or
“broadband” services are any services that operate at more than 256
to 384 kbps, though purists would argue that broadband must truly be
more than T1’s 1.544 Mbps rate.
Where some
confusion arrives is in the mission of the concept, and that lets us
divide the market into two very specific segments—fixed broadband
wireless to serve non-mobile sites, and mobile broadband wireless to
serve portable/movable/moving locations.
The mobile broadband market would be best characterized as the
“3G and beyond” space, which we’ll cover in a later issue.
At this time, we’ll also cover the “portable” wireless LAN
applications in the carrier space. For
now, we’d like to say that we believe that mobile broadband
applications (like Nortel’s video-PDA-terminal commercial) are a long
way in the future, and that fixed broadband services would have to be
pervasive first in order to develop a mobile market.
The “wireless LAN loop” application, while it has promise,
may suffer from the fact that it uses unlicensed spectrum and is thus
susceptible to interference from the same, or other, applications
provided by other carriers or even users.
So, it’s fixed
broadband wireless that we’re going to shoot for.
This type of technology can be subdivided too, into
“point-to-point” (PP) and “point/multipoint” (PM).
The former would consist of applications of wireless where a pair
of stations created a single customer link, and the latter where some
form of multi-user arbitration permitted a single base station to serve
multiple customers.
PP broadband
wireless is clearly a specialized application.
The reason is that the cost of RF stations isn’t trivial, and
having two dedicated to a given link guarantees that the capital
investment in the connection will be significant—up to tens of
thousands of dollars, particularly if installation is included.
This would mean the access cost to the customer would have to be
in the thousands of dollars, making the technology too expensive for all
but larger business applications. PP
wireless, in fact, is already in use for microwave transmission of
information within companies, and also by some carriers in distributing
access bandwidth to high-value customers.
It’s PM broadband
wireless that has aroused the most interest.
There are two primary versions of PM broadband wireless;
Multichannel Multipoint Distribution Service (MMDS) and Local Multipoint
Distribution Service (LMDS). Both
are currently offered by at least some players, but most believe that
LMDS is the future direction the market will take—if it takes any
direction. We’re not so
sure.
All PM broadband is
based on a system of cable-modem-like frequency or slot allocation, so a
single base station can serve multiple clients at the same time.
This allows the cost of one side of the wireless connection to be
shared across many clients, reducing overall capital cost.
CDMA, TDMA, or virtually any kind of –MA will work to provide
for bandwidth distribution among the users.
The concept is really quite close to that of cable data systems,
so much so that some call it “wireless cable”.
Another kind of
“sharing” permitted by the PM broadband systems is frequency re-use.
If the broadband service range from a given base station is
relatively short, the same frequency can be used in several places in
the same general geographic area as long as no client is in range of two
different base stations operating on a single frequency.
The service area or “cell” size is dependent in part on the
frequency; MMDS has a longer range and thus does not allow as much
re-use. On the other hand, a
short range or small cell size requires greater multiplicity of base
stations, which increases cost.
Frequency plays a
role in other ways. Higher
frequencies carry more data per unit of spectrum space, so LMDS has a
larger capacity than MMDS for a given base station.
With LMDS, for each
allocated frequency, a series of 50 channels, each supporting about 40
MHz traffic rates, is created in the downstream-to-the-user direction,
with smaller and slower uplinks. Typically,
the downstream direction is further multiplexed by customer to break the
capacity up into more usable chunks.
The whole thing operates as a kind of shared RF LAN based on a
micro-cell concept similar to that used by PCS.
MMDS works at lower frequencies (about 2 to 3 GHz) and LMDS at a
high (28 to 40 GHz) frequency. Both
licensed and unlicensed frequencies are available, and while the MMDS
strategy offers more distance and somewhat better immunity to
obstructions. MMDS, with
longer range, is probably better for more rural areas, but the longer
range will also limit frequency re-use and reduce the potential number
of customers that a single frequency license can serve.
It also presents lower data capacity and seems likely to be
gradually displaced. We
should note that both technologies are considered line-of-site, which
means there has to be a visual path from base to client antenna.
Because LMDS is
prone to interference from foliage and even atmospherics, generating an
economical cell size can be challenging in some service areas.
To counter this, LMDS provides two different distribution
architectures, a single central point (“one-layer”) or a hierarchy
of antennas; metro and more local (“two-layer”.
Two-tier systems have a larger cell size, and the second tier is
created using set of repeaters less costly than the base station, so
capital cost doesn’t multiply in a linear way with cell size.
The cost of PM
broadband is made up of both recurring and one-time or capital
components. The capital
costs include the cost of the license (which will vary depending on
where the service area is found—it’s typically an auction process),
the cost of the base station(s), and the cost of the client equipment.
Recurring costs include the usual administrative costs, plus
craft costs associated with maintaining the transmission facilities and
dealing with any customer problems.
The big advantage
of this sort of approach, of course, is that it doesn’t require any
wire-line infrastructure be installed or acquired.
A single LMDS cell can cover an area of from 1 to 2 miles radius
in two-layer mode, with a series of repeaters serving small residential
or business zones and fed from a common central point.
That service area, which could be as large as 12 square miles and
enclose literally millions of prospects is “on the table” for the
cost of the license, base station, and repeaters.
Each customer then adds incremental capital equipment as they
sign on.
Of course, a single
LMDS cell can’t possibly serve a million users with any grade of
service. How many users
could actually be supported is somewhat dependent on the expected grade
of service. If we assume
that the subscriber’s peak access rate is 1 Mbps (consistent with
competitive wire-line technology) and that oversubscription of 15:1 is
permitted, the LMDS microcell could support 50 channel slots of about
600 users each, or 30,000 customers.
High LMDS license
fees, resulting from auction speculation, contribute to this problem.
If a very high license fee is paid, recovering cost could be
challenging even with high customer counts.
And, of course, Rome wasn’t built in a day, and a customer base
of 30,000 within a 12-square-mile area could require some marketing time
and cost. High license
charges also tend to front-load LMDS investment, and this erodes the
advantage of wireless in eliminating dependence on fixed infrastructure.
With MMDS, we can
expect at least a 20-mile radius on the service area, providing there
were minimal obstructions. MMDS
has lower data rates and less ability to multiply frequencies, so it
also serves a smaller customer base within a given geography.
We estimate an MMDS system can support no more than about 10,000
customers from a single frequency in a given base cell, and probably
less. But MMDS licenses are
usually cheaper, and MMDS capital equipment is likewise less expensive.
MMDS is a rural
winner, in our view, and a strong contender in upscale suburban areas.
It may also be a winner in areas where high tower locations (on
mountaintops, for example) could create a very large service area for a
single cell. Since the
service geography for MMDS is large, it is probably easier to develop a
starting MMDS footprint in a given metro area using conventional
advertising; a single cell might serve the entire area.
Expanded customer needs might then drive deployment of additional
capacity. With LMDS in the
same situation, getting a service footprint as large as the
advertising/marketing range would require a much larger number of
initial cells, and a higher investment.
LMDS is the
traditional urban pick, providing that buildings and other obstructions
don’t restrict that a full 12 square mile cell size. But
LMDS’ small service footprint per base station could make it hard to
market an LMDS service unless the carrier bites the bullet and deploys
LMDS over an entire metro area, with many cells and a higher startup
cost. We’re also concerned
about the competition that could be offered in urban areas, competition
that could lower rates of return for the LMDS carrier.
The question is
whether either of these can generate enough ROI to make the players
survivors, particularly in the first two years.
We think, based on current license trends and Wall Street
expectations, that the chances of a near-term success are small.
Clearly, the residential market is too price-limited for early
success with any fixed wireless technology.
Unfortunately, the business market may be too quality-sensitive.
Service problems with fixed wireless may not be legendary, but
the risk of such problems have always been perceived to be three to five
times higher than those of wire-line infrastructure (according to our
research in 1998 and 1999, for example, buyers believed outages in
wireless would occur at over five times the rate of standard loop
technology).
Running our usual
magic ROI calculations on the wireless market shows to our satisfaction
that no fixed wireless option can possibly generate enough ROI to make
the carrier survivable, unless we assume that the public willingness to
pay for fast Internet access increases dramatically, and that neither
the cable players nor the ILECs are able to develop wire-line
approaches. Obviously, none
of those three variables are likely to move in the direction favorable
to wireless.
Our scenario for
fixed wireless is one of specialized use.
We think that the ILECs will deploy it to meet the requirements
Congress and the FCC are likely to impose on uniform empowerment of
rural communities. We think
that some specialized CLECs may also be able to penetrate that same
market, providing that the tax credit legislation now before Congress to
help rural deployment actually becomes law.
With wireless, as
with other technologies, the classical wisdom is that it’s a
competitive technology to DSL, and thus will deploy primarily if DSL
continues to lag. We believe
the opposite is true. DSL is
needed to create a large broadband community, and we believe it alone
can do that. Once such a
community is established, there will be segments of the
residential/business market that cannot be served optimally with DSL.
Wireless and other technologies will serve these segments. It
won’t make zillions of bucks, and most of the current license-holders
may sell off their licenses before anything happens.
We’ve had other,
earlier, wireless data technologies.
Remember LDDS? None
of them have played a significant role in developing networking’s
future. We think that while
the current technologies have promise, the inability of our marketplace
to face reality is creating a set of conditions under which those
promises won’t be easily realized.
This section of our newsletter is
dedicated to the service management framework response of new-age
service management player Mantra. Sorry, but it's
for subscribers only!
Netwatcher is a copyrighted information product of CIMI Corporation. Subscription is on a controlled circulation basis only. Subscription information may be obtained by clicking here.