This is from NTRO's VP of marketing for Netro wrote this article. Wow. She takes the high road too,
and doesn't pump her own company directly. This article gives a great overview of
the different access methods out there, where wireless fits in, and a little about how it
works. I'm impressed by her style, and the style of a company which would hire a
marketing VP capable of this type of work. The article is a reprint from Wireless
Review....
wirelessreview.com
By Cynthia Hillery
The metropolitan access market (small- to
mid-size businesses with up to 1,000 employees)
is bandwidth deficient. Historically, this segment
is overcharged and under-served. Traditional
ILEC voice-centric solutions cannot
accommodate its growing need for data services
such as in-house e-commerce management
applications, including orders and billing for retail
sites and network applications.
With a complex mix of voice and data emerging
— Pioneer Consulting reports that data per user
will increase 5-fold over the next five years (see
Figure 1) — the market has opened for CLECs to
compete with incumbents through value-added
services and integrated solutions.
Figure 1. Pioneer Consulting reports a significant demand for
data.
To compete for small medium enterprise (SME)
revenue, CLECs will have to provide a last-mile
solution that encompasses three critical areas:
bandwidth flexibility to accommodate bursty data
traffic; service integration to combine voice and
data within a reliable system; and coverage
economics for quick market entrance and
incremental growth.
New technology is the only way CLECs can
begin to compete in the incumbent's backyard.
Many different access technologies can address
the last-mile bottleneck: fiber, xDSL, cable
modems, 3G, broadband satellite and broadband
fixed wireless. Each has its own strengths and
weaknesses, and caters to the needs of different
market segments. Many industry experts believe
the technology with the best fit for the SME,
particularly larger customers in the segment, is
point-to-multipoint broadband wireless access
(BWA). Its incremental growth potential and
ability to integrate and provision value-added
services could determine its success.
Fiber
There is no doubt that fiber-optic backbone
provides the fastest speeds, the most bandwidth
flexibility, and a high degree of reliability and
security, but also it has disadvantages. For
example, fiber is connected to only 3% of all U.S.
commercial buildings. Although carriers have
deployed fiber effectively between high-capacity
switches and high-traffic nodes in
communications networks, deploying fiber to the
residential and business customer is capital
intensive and often cost prohibitive ($200,000 to
$300,000 per mile), especially in non-urban
markets. Few markets ensure the revenue to
justify this expense.
Another disadvantage of fiber is the significant
cost of installing fiber routes without assurance
that customers will use the built-in capacity fully.
Carriers have to estimate customers' needs along
proposed fiber routes, and often costly
adjustments are required after installation,
depending on the coverage or shortfall of
customer demand. Therefore, the telecom
industry generally has used fiber to provide
high-value and premium-priced services to large
customers with business-critical applications.
Connecting the metropolitan access market via
fiber is time-consuming and risky. Because
carriers typically don't offer fiber to SMEs, there
has been no feasible technology to provide them
with high-speed data and voice services.
XDSL
Thanks to successful IPOs of companies such as
Covad and Rhythms, xDSL technology has
become a hot topic lately. The xDSL technology
family concentrates on maximizing the
throughput of a telephone company's existing
twisted-pair copper network. The xDSL variants
take different approaches to transmitting the
maximum number of bits per second through the
legacy copper access network. The throughput
on a twisted copper pair is a trade-off between
bandwidth and distance: The greater the
bandwidth, the shorter the network reach.
All data transfer rates, however, depend on the
copper plant's condition. Copper lines currently
installed have not met initial expectations for
xDSL applications. Usually, xDSL enables
carriers to leverage existing copper because there
is no need to lay new infrastructure such as a
second pair of copper wires as is the case with
ISDN.
XDSL is a good technology for transporting data
across the metropolitan access market but has a
limited ability to converge voice and data. Its true
niche is as a data-transport technology for the
small office/home office (SOHO) market, where it
will compete directly with cable modems.
Cable Modems
Cable modems and hybrid fiber coax provide
faster speeds than analog modems or ISDN. This
technology is the originator of the recent AT&T
activity in the cable market (TCI and MediaOne).
Cable operators are entering the CLEC market
rapidly.
A coaxial cable has more bandwidth than twisted
copper pair, but most companies built cable
networks with limited 2-way capabilities. The
traditional coaxial networks are 1-way networks,
which means they transmit signals only from the
cable company to the subscriber. The cost to
upgrade cable networks, which are largely
analog, to a digital format and to provide
switched services, parallels the cost required to
install fiber.
Specifically, for small- and medium-size
businesses, there are drawbacks to using cable
modems for voice and data services. Cable
modems are asymmetrical, so they do not fit well
with future e-commerce applications. Cable
network upgrades are slow with substantial fixed
costs. Moreover, cable companies have a
reputation of providing poor service to consumers
with long outages.
Security is another concern. Cable modems
operate over a shared system, increasing risks
against other technologies. Cable-system
powering requirements increase the risk of
relying on this technology as the sole means of
communications. Typically, metropolitan-access
businesses strive for telecom services that are
compliant with 99.999% reliability. It still is
uncertain whether cable operators will be able to
deliver this reliability.
Cable modems are mainly a residential, or SOHO
play, and will emerge as an impressive competitor
against ILECs and xDSL technology.
Wireless Arena
Broadband satellite and 3G technologies have
been field tested, but not proved, and lack defined
standards. Although development continues,
deployments will be limited for quite some time.
They promise to provide high-speed access for
mobility, but a business-quality launch is several
years away.
That leaves BWA solutions as the best choice for
the metropolitan access market from an economic
and ease-of-deployment standpoint. Today's
point-to-point technology will serve as a
backhaul option. Point-to-multipoint is the
well-designed technology that will facilitate
multiple-subscriber access to the network for
integrated voice and data applications. It can
address the three main concerns of the
metropolitan access market: bandwidth flexibility,
service integration and coverage economics.
Point-to-multipoint deployments will accelerate
by year-end. Currently, fewer than 12 networks
are “on-net,” but carriers are testing roll-out
plans and network integration. (See Table 1.) As
a general guideline, three months after testing,
carriers move to citywide deployment; six to nine
months later, they deploy nationwide. The
Strategis Group estimates the rapid growth of this
market will facilitate
customer-premise-equipment (CPE) revenues
close to $3 billion by 2003. Networking
equipment will equal, or at least double, the CPE
investment over the same time period.
ART
The company has connected 126
buildings to its network with service
orders from 250 customers. ART
offers commercial service in Portland,
OR; Phoenix and Seattle via
point-to-point technology. In addition,
it is trialing point-to-multipoint
throughout the country.
Formus
It is conducting point-to-point testing
and trials in Budapest; Strasbourg;
France; and Denver. The company
will introduce live point-to-multipoint
service beginning this summer and
throughout 1999.
Nextlink
The company is testing and trialing
point-to-point and point-to-multipoint
technology in its Plano, TX, lab
facilities and commercial
environments in the United States. It
will offer commercial broadband
wireless service in 1999.
Teligent
It is offering commercial service in 27
markets comprising 430 cities and
towns. It has more than 3,100
buildings under lease or option. The
company has equipped 444 buildings
with fixed-wireless installations using
point-to-point and point-to-multipoint
technology. By year-end it expects to
be in more than 40 markets.
WinStar
The company has 17,000 customers
and access rights to 4,800 buildings.
Currently, it offers commercial service
in more than 30 markets using
point-to-point and point-to-multipoint
technology. It plans to reach 45 U.S.
markets as well as six international
markets by year-end.
Table 1. Deployment progress and plans for some broadband
wireless carriers
Point-to-multipoint is a complex networking
technology that has a high degree of reliability
and scaleability, along with minimal infrastructure
and real estate requirements, while providing
faster speeds than cable modems and xDSL
technologies. Broadband wireless technology is
optimized for business customers requiring
integrated voice and data capabilities within the
2- to 8-T1 range.
In addition, this new technology still is
undergoing significant developments. For
example, there are two key design variances for
point-to-multipoint systems.
One difference is in the transmit/receive signal
duplexing mechanism: frequency division
duplexing (FDD) vs. time division duplexing
(TDD).
FDD divides the upstream and downstream
transmission channels into separate frequencies
within a sector, so each radio transmits
simultaneously and continuously. TDD divides a
single channel for both upstream and downstream
transmissions by time instead of frequency, so for
every given millisecond, some time is dedicated to
transmitting and some to receiving.
FDD is commercially viable and TDD is in
development. For international growth
opportunities, FDD is the only compliant design
for the ETSI standard.
Another design variance is in the spectrum
resource access mechanism: FDMA vs. TDMA.
FDMA divides the spectrum into separate, unique
bands. Each user acquires a band, or pipe, to
access the network. But this does not allow for
dynamic bandwidth allocation, and much of the
system's capacity may remain idle. First
generation point-to-multipoint products were
FDMA-based, but the market demand for service
integration, bandwidth flexibility and economic
growth has prompted a second-generation
evolution to a TDMA architecture.
TDMA users access the network on a
time-shared basis, where each slot is directed to a
fixed subscriber — a packet in every time slot.
With a media-access-control protocol that
enables multiple subscribers to burst with their
data transmissions, TDMA facilitates dynamic
bandwidth allocation.
TDMA requires ATM technology to maintain
voice quality of service and prioritize subscribers
for the best spectrum management. It appears
TDMA, coupled with ATM, is becoming an
industry standard rapidly.
The most technologically advanced
point-to-multipoint systems transmit voice, IP and
frame-relay traffic to and from multiple
subscribers. A well-designed network with a
TDMA/ATM architecture can perform statistical
multiplexing and dynamic bandwidth allocation
to reduce the cost per DS-0 drastically.
Compared to fiber at an average of $1,513/DS-0,
point-to-multipoint on a fully loaded base station
is $175/DS-0 to $400/DS-0.
Market Divergence
With emerging broadband technologies, the
market diverges between the residential/SOHO
market, the metropolitan access market and large
enterprises. There is a broadband access solution
that is tailored to every subset, but the
metropolitan access market comprises the
greatest percentage. Cable modems and xDSL
will dominate the market space demanding less
broadband access capacity. Fiber, enabling
expansive broadband requirements, will reach its
saturation level because of fixed labor costs.
These technologies leave a prominent access gap
for small- and medium-size businesses.
Point-to-multipoint could emerge as the solution
for closing the gap between the customer premise
and carrier backbone.
Hillery (cindyh@netro-corp.com) is Netro vice president of
marketing.
Technology Basics
Point-to-multipoint systems divide the air
surrounding a tall building in which a hub/base
station resides into multiple pie-shaped sectors.
(See Figure 2.) The range of a system is 3km to
15km, depending on frequencies and weather
conditions. The base station aggregates voice and
data traffic from all system subscribers and is
connected to the central office by fiber or
point-to-point radio connections.
Figure 2. A basic point-to-multipoint design
The system connects with several T1 ports as
well as Ethernet and frame-relay ports. The
subscriber-access system transforms integrated
voice and data communications to an
intermediate RF and sends it to an outdoor
rooftop radio unit. Information is converted into
the appropriate band and transmitted to the
base-station radio. A modem at the base station
demodulates the signal into digital stream and
multiplexes it through the base station onto the
ATM network.
The subscriber radio unit is the size of a shoebox,
and you can mount it on a building or behind
glass. Unlike the fiber set up, which owes
approximately 80% of its cost to labor, the
point-to-multipoint system installation comprises
less than 20% of the system cost and only takes
hours. |
