Not sure what the question is, Ray. Please expand a bit.
While we're on the topic of pulling fiber overseas, this month's Lightwave Magazine's supplement, Fiber Exchange, has a very good story on European Fiber buildouts. This is a real eye opener, IMO.
FiberExchange is at:
fiber-exchange.com?Section=External&Subsection=/
The article, BANDWIDTH WITHOUT BORDERS: The Race to Interconnect Europe, is at:
pennwell.shore.net
The most impressive graphic is the last chart in the article, which shows each of the carriers network sizes and types of fiber used. You will note that the oldest PTT carriers, such as France Telecom and DTK only have 2 strands lit, and they are among the oldest in place. The newer carriers, in contrast, have up to several hundred strands placed each. Quite interesting. The text portion is copied below for posterity. Go to the url, however, for the graphics. Enjoy.
FAC
=========
BANDWIDTH WITHOUT BORDERS:
The Race to Interconnect Europe
Incumbent and emerging operators are rushing to build-out fiber
networks in Europe as government-controlled monopolies lose their
stranglehold on long distance and data traffic soars.
BY KATHLEEN RICHARDS
Fiber-based networks interconnecting
cities and countries through out
Europe may soon transport voice and
data, in, out, and around the
continent, and allow customers to buy
services from a single pan-European
carrier. This "borderless"
communications infrastructure is being
fueled by the same factors driving the
fiber-network frenzy in North
America-liberalization, market
deregulation, Internet growth, rapid
increases in data traffic, and willing
investors ready to pour private capital
into fiber-based networks, carriers,
and equipment manufacturers.
As of Jan. 1, 1998, liberalization took
effect in many European
countries-Austria, Belgium, Denmark, Finland, France, Germany, Greece,
Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden, and the
United Kingdom. In a broad sense, it opened up the international
(long-distance) markets controlled by government-owned monopolies (postal,
telegraph, and telephone companies, or PTTs) to facilities-based competition.
The incumbents would have to compete with other operators on price and
technology and negotiate interconnection agreements to provide local circuits
and terminate international traffic.
"With liberalization, there are some restrictions," says Patrick Fay, analyst at
market researcher KMI Corp. (Newport, RI). "Operators still go through some
hoops with the telecom authorities to get licensing in some countries, but
essentially, now there is a way to compete with the incumbent carriers."
Prior to liberalization, only a few carriers were building pan-European networks,
observes Fay. The first was Hermes RailTel, which announced a network in
1996, followed by Esprit Telecom, and MCI WorldCom, which announced its
Ulysses network in 1997. Hermes RailTel and Esprit Telecom have since been
acquired by Global TeleSystems (GTS), headquartered in Washington, DC and
London.
Crowded continent
Liberalization was designed to promote international competition, and it has. At
least 21 fiber-based operators are developing intercity backbones that range in
size from Eurotunnel Telecom's (ETT's) 320-route-km Euro Tunnel to Interoute's
20,900-route-km I-21 network. Many of these operators have focused on Tier 1
routes across France, Germany, the United Kingdom, and the Netherlands,
countries that account for 63% of the current traffic in Europe, according to
market researcher Ovum Inc. (United Kingdom). That traffic is largely data
generated by Europe's four major financial centers-Amsterdam, Frankfurt,
London, and Paris. Consequently, the highest bandwidth demand is in
northwestern Europe.
Like the United States, the European market is poised for exponential growth
in traffic and capacity requirements. "There's been a frenzy of building activity
in the last 12 to 18 months," says Barry Flanigan, senior consultant at Ovum.
"And that's in anticipation of future demand when this huge data wave really
takes off in Europe and creates the traffic to fill up this capacity. But that's still
in its early stages."
By 2005, total pan-European bandwidth demand will increase to more
than 9 Tbits/sec from 339 Gbits/sec this year. Cumu-lative traffic
volume is expected to reach 2.366 Tbits/sec from 72 Gbits/sec during
the same period.
Total pan-European traffic volume is forecast to almost double each year during
the next five years, reaching 2.366 Tbits/sec in 2005. During the same period,
intra-European traffic will climb from 24 Gbits/sec to 767 Gbits/sec. Ovum's
annual traffic-volume forecasts through 2005, segmented by where the
pan-European network traffic originates and terminates are shown in Table 1.
During the same five-year period, total pan-European bandwidth demand is
forecast to increase to more than 9 Tbits/sec in 2005, from 339 Gbits/sec this
year, according to Ovum (see Figure).
Handful of cities
By 2001, at least 135 European cities will have connections to pan-European
networks. The largest markets will be Germany and the United Kingdom (28
cities), followed by France (15 cities) and Belgium (12 cities). Most countries,
however, will have "only a handful of cities" served, according to KMI research.
About 16.7 million fiber-km will be installed from early 1998 through 2002,
according to KMI's pan-European forecasts. Interoute's I-21 network,
announced last June, represents almost half of this volume. New operators will
account for 90% or 15 million fiber-km, while incumbents will install 10% or 1.7
million fiber-km. Nonzero dispersion-shifted fiber (ITU G.655) will represent
76%, or 12.6 million fiber-km, of the fiber installed, according to KMI.
Conventional singlemode fiber will account for 24%, roughly 4.1 million
fiber-km.
"The nonzero dispersion-shifted fiber [NZ-DSF] allows the carriers to forego
building regeneration sites that, with conventional singlemode fiber, have to be
installed at more regular intervals," says KMI analyst Fay. Building
regeneration sites involves permits and other regulatory hurdles. "So a lot of
nitty-gritty construction details are avoided," observes Fay. The nonzero
dispersion-shifted fiber is also optimal for higher bandwidth technologies such
as dense wavelength-division multiplexing (DWDM).
In addition to adopting NZ-DSF, another trend emerging among the fiber-based
pan-European operators is higher fiber counts per cable, according to KMI
research. MCI Worldcom, which announced its network in March 1997 is using
a 48-fiber cable, while Metromedia Fiber Network, which announced its network
in August 1998, is installing a 216-fiber backbone (see Table 2).
"The strategies of the various pan-European players vary," says Ovum's
Flanigan. "Some carriers are putting in what I would call fiber-rich networks,
lots of fiber pairs and dark space. And then at the other end of the extreme,
you've got some carriers building relatively fiber-thin networks with a small
number of fiber pairs and using lots of DWDM to get the capacity that they
need."
Operators are also digging duct space as a way to futureproof their networks,
according to Flanigan. Duct space can accommodate future generations of
fiber, which may, for example, handle higher channel counts and DWDM
equipment. It also serves as a valuable asset to sell or swap with other
carriers.
"Chances are, carriers won't be building all of the infrastructure themselves,"
says Flanigan. "They'll install the fiber where it is economical to do so, but in
some regions, they will lease capacity or infrastructure from carriers or other
organizations that are already there." This strategy is especially likely to occur
as carriers try to expand into southern, central, or eastern Europe in the next
12 to 24 months, where in some areas it may be difficult to build on the terrain
and to obtain rights-of-way. This type of expansion presents an opportunity for
niche operators to lay the infrastructure, then sell capacity to other operators
trying to enter the region. Iaxis, for example, has announced plans to build a
submarine cable in the Mediterranean Sea and sell capacity to other carriers.
In addition to the long-haul backbone activity, network installations in
metropolitan areas are on the rise. According to KMI research, the number of
pan-European metropolitan area networks (MANs) reached 55 in 1999, up from
20 in 1998 and 12 in 1997. As operators build-out their intercity backbones and
start to develop a customer base, many want to offer customers an end-to-end
solution. The next step is to acquire or build city networks. Early entrants MCI
WorldCom and COLT Telecom were joined in 1999 by new metro players,
including Metromedia Fiber Network, Level 3 Communications, and GTS.
Unbundling the access
Despite market competition among intercity network operators in many
regions, incumbents still control the local loop and keep prices artificially high.
"Where the price bottleneck is occurring now is in the access market, much
like the situation you've got in the United States," says Flanigan. "You see
prices in the backbone tumbling down, but in the access market, it is a lot less
competitive. The national incumbents still have a stranglehold on what is going
on there. So you get in a strange situation, where in some cases it is cheaper
to lease a long-distance circuit-say the length of the United Kingdom-than it
would be to lease a short-distance circuit to get from the point-of-presence out
to the business premises."
Regulators in Europe are focusing on the access market, and the European
commission has issued recommendations in the last few months to try to
control access prices. "That's very important to what happens in the
backbone," says Flanigan, "because all this cheap bandwidth is not going to
filter through to the end user, if the prices in the access market stay very high.
It is a potential barrier to the uptick of high-bandwidth applications and the
whole growth of e-commerce."
Is the market saturated?
While analysts expect to see more consolidation and shakeouts among
operators, as well as new carriers and service providers entering the market as
the cost of pan-European capacity drops, it is unlikely that additional
fiber-based carriers will emerge in the next few years.
Click image for large table
"We believe there is a fiber shortage now for some crucial areas, and that is in
the United States and in Europe," asserts KMI's Fay. "It would be difficult for
new carriers to come along and make grand announcements about installing
all this fiber, because they're not even in a position to begin installing it until
maybe 2002."
The unexpected demand for NZ-DSF in the United States and Europe, which is
provided by and large by Corning and Lucent, has created a bottleneck for
manufacturers. "It's also having the effect if you want conventional singlemode
fiber that's become more difficult to obtain," says Fay. "Corning and Lucent
have had to fly in fiber from Japan to meet their conventional singlemode
commitments."
Pan-European competition will continue to heat up, however, as bandwidth
pricing drops and the newer technology allows operators to continue to lower
the cost per bit and undercut the cost of transport on incumbents' legacy
networks. Cross-border tariffs have already plummeted on some high-traffic
routes, such as London to Paris, where bandwidth pricing dropped by about
40% to 50% during a six-month period in 1999, according to Ovum. "A lot of
capacity is going in in Europe, and one of the consequences is that bandwidth
prices are dropping rapidly," says Flanigan. "I think you will start to see a lot of
service providers and carriers that aren't in Europe coming into the market and
taking advantage of that cheap capacity." |
