From Telephony Magazine
Bandwidth
September 14, 1998
Multiservice moves in
SUSAN BIAGI
Sell the network over and over. This mantra defines the
purpose of multiservice access devices. These
devices--primarily concentrators and switches--are gaining
popularity as service providers seek to expand their service
offerings.Why sell just voice if you can also sell frame
relay? Or Internet connectivity? Or digital subscriber line
(DSL) technology? Or, down the line, virtual private
networks (VPNs)?
Access equipment vendors strive to ensure that carriers
make the most of their networks. Cisco Systems and
Northern Telecom are key players in multiservice access.
Ascend keeps adding slot cards that enable its MAX
products to support new services. Lucent Technologies
acquired access equipment vendor Yurie Systems five
months ago, upping its commitment to multiservice access.
Recognizing the market potential, a spate of start-ups is
stepping into the multiservice space. These new vendors
are all over the map: Some focus on specific protocols or
network technologies. Others, such as Advanced Switching
Communications, propose the idea of a gateway, through
which all traffic will pass and be forwarded intelligently.
The trend in the multiservice access arena appears to be
"the more, the merrier." But how do products differ? And
how are vendors addressing the different needs of
competitive and incumbent local exchange carriers?
Lucent's Access Concentrator is an asynchronous transfer
mode switch to which carriers add cards. The cards
convert legacy or protocol traffic, such as frame relay, to
ATM and transport the data across the ATM network.
Although frame relay and ATM register huge growth
markets, time division multiplexing (TDM) is still growing,
says David Haas, director of marketing for Lucent's access
business.
Ascend puts its MAX TNT product in this category,
although some industry watchers claim the 9000 product
family is a better fit. The TNT supports frame switching,
basic-rate ISDN, DSL, and voice-over-IP connections
simultaneously. It combines Layer 2 switching and Layer 3
routing functions, bringing them into the WAN access
space. This functionality opens the doors for providers to
offer VPN and port wholesaling services in the future.
The ultimate goal is to support the "new public network,"
where all types of traffic mingle on one network and quality
of service is maintained, says Kurt Bauer, Ascend's vice
president of access product management.
The TNT is a step in that direction. It handles ATM, frame
relay, IP, faxing and Internet tunneling services, as well as
conventional remote access, on a common infrastructure,
says Ted Butch, director of access switching product
marketing for Ascend. "The software you load gives it the
character you're looking for."
The big vendors may be putting their weight behind the
multiservice access concentrators, but smaller vendors are
fueling the market's rise, says Frank Dzubeck, president of
Communications Network Architects. "I see it happening
this year or in the first quarter of next year. It's coming
about from a number of start-up companies with more
robust products."
One example is Advanced Switching Communications.
This new company is developing "multiservice gateways,"
says Ron Westernik, the company's marketing vice
president. He contends that traditional concentrators need
to get smart to survive and a new product category will
emerge to reflect that need.
The new breed of device will incorporate several specific
functions, Westernik explains. It has to manage
downstream elements, such as a DSL access multiplexer, a
server, a modem bank or terminal services. It must have
multiple level connection analysis capabilities and traffic
shaping functions. And he says, "It has to have multiple
personalities." By that, Westernik means the device has to
be flexible, adaptive and economical while switching or
inverse multiplexing any number of access trunks to any
number of network trunks. Protocol-independence is also
necessary along with an inexpensive price tag.
Can it be done? Westernik thinks so. But it requires a new
device architecture. Adapting current concentrations won't
suffice, he says. He unveils the six components that
embody this new architecture: distributed power, a
switching fabric that is aware of cells and frames, slot
independence, universal protocol independence,
carrier-class synchronization and multilayer awareness.
"A distributed architecture offers the best
price/performance [ratio] at the lowest cost of entry," he
adds. "We're building a product around the things I've
described. It's a multiservice gateway. It's not a switch or a
concentrator. It's where those [device vendors] have to go
with something if they intend to get into this space."
And getting into this space they are. One reason the market
is teeming with new entrants is because CLECs present a
great service opportunity. CLECs and Internet service
providers are competing with incumbents on price, so they
need to offer an array of services and incur little overhead,
says Cathy Gadecki, a director at TeleChoice.
With no legacy systems, competitive service providers are
designing multiservice networks from scratch and avoiding
costly overlay networks. Maintaining multiple networks is
expensive in terms of hardware and operations, and they
don't scale well. One-device solutions fit the bill.
"Vendors are combining some platforms that allow [service
providers] to roll out multiple services," Gadecki says.
"New carriers try to justify the [equipment] expense with
one service and readily add new services to their portfolio
with new cards. They are incrementally building on the
same infrastructure from an equipment, process, people
and tools perspective. It's really attractive to the new
providers," she says .
The
multiservice
network
allows
CLECs
to
offer
customers
one-stop
shopping,
even
if
the
carrier
doesn't
know
what
services
will
generate demand. The key is bringing services out to the
edge, says Lucent's Haas.
Instead of placing a time division multiplexer there, carriers
"are using an ATM multiservice access concentrator at the
edge of the Sonet ring," he says. "That puts them in a
position to offer frame relay, ATM or TDM [depending on
end user requests]. They might be 70% or 80% sure, but
they can generate a lot of business if they are more flexible
depending on what services end users want."
It can cut costs, too, Gadecki says. For example,
interexchange carriers that deploy multiservice access
devices close to the customer improve reliability and cut
costs associated with backhauling .
Supporting
services
at
the
edge
opens
up
the
small
office/home
office
market.
Seeking
to
eliminate
the complexity of branch office routers, these customers
want to connect via ATM, Haas says. "We are replacing a
lot of small office routers. They are cutting out the
[multiplexer]. They want to put a multiservice access device
at the edge to aggregate all the traffic."
Incumbents have the challenge of getting devices that
interoperate with existing equipment and meet their existing
service definitions. The long decision-cycle process is
fortuitous for CLECs, however. They can deftly offer
bundled services, grabbing marketshare before incumbents
have a chance to deploy additional services.
The driving force is that carriers are preparing for the
onslaught of data traffic. Data doesn't flow over the
network the same way voice does, and entrepreneurial
companies will exploit that. They will integrate more
services while driving down costs.
Although several start-ups haven't released product or
pricing, a large switch can cost a bundle--far more than a
new CLEC can afford. The onus is upon vendors to offer
flexible, low-cost devices that CLECs can cost-justify and
deploy quickly. That will happen over the next few months,
Dzubeck says, and multiservice networks will take hold.
"This business is going to change," he says. "These
[start-ups] are all coming from the data world, and these
guys are going to rearrange the port costs."
Susan Biagi is Switching & Transmission Editor for
Telephony magazine. |
