What's really behind the Ascend
and Cisco deals
By Thomas Nolle
Network World, 09/07/98
On the surface, the news seems simple. Ascend buys
Stratus. Cisco buys Summa Four. IP giants battle for
positioning in the holy war of convergence, or voice over
IP, or whatever. But nothing is what it seems to be on the
surface these days.
Look at two other news stories. Lucent gets orders from
regional Bell operating company SBC and competitive local
exchange carrier (CLEC) Dakota Telcom for billions of
dollars worth of 5ESS switches. If we're converging on a
new IP-based voice infrastructure, the old technology is sure
getting a heck of a send-off.
Now it's possible that Cisco and Ascend planners have their
heads in the converging clouds, so to speak, and have
missed the other news stories. It's also possible that the first
two news items were really caused by the second two. If
Class 5 voice switches are still worth billions, maybe IP
players need to get into that market. (Note, I'm not saying,
"Replace that market with IP," but rather "get into that
market.")
There's an architecture - a set of standards - that Cisco and
Ascend could use to threaten Lucent and Nortel (not to
mention Siemens, Ericsson and others) in the Class 5 switch
market. It's called advanced intelligent network (AIN).
AIN is a Bellcore initiative to restructure the way voice
telephony works to make it look more like client/server
computing. In traditional voice networks, all service features
had to be programmed into the big central office edge
switches, or Class 5 switches. The software was so
complex that features were planned two years in advance,
and the cost of the devices kept mounting. AIN is targeted at
fixing that problem. In an AIN, service intelligence is
distributed among service control points (SCP), service
switching points (SSP) and intelligent peripherals. These
smart gadgets are linked in a control network using signal
transfer points (STP). What runs that control network?
Signaling System 7 (SS7). What's the common thread in the
Cisco/Summa Four and Ascend/Stratus deals? SS7. Summa
Four is big in the intelligent peripherals market; Stratus is
big in intelligent peripherals and STPs. Get the picture?
If AIN were adopted, it would make today's Class 5
switches a distributed series of switches and computers
linked via SS7. The switches would be SSPs, not much
more than a dumb matrix that nearly anyone could build. All
the service features - the SCPs and intelligent peripherals -
would reside in the computers.
AIN was designed to make it possible to deploy new service
features faster. It was also designed to reduce costs to
carriers and open the market to more competition. Those are
goals competitive carriers would value, yet a large local
exchange carrier (SBC) and a CLEC (Dakota Telcom) just
bought traditional Class 5 switches from Lucent. Could they
have done so in part because no organized, AIN-like product
set is available? Today even small switches are built on
monolithic lines, without the concepts of distributed
computing that AIN has represented for nearly a decade.
AIN could take off, though, if somebody brought the right
combination of computer and switch technology together
and linked it with SS7. By developing a Class 5 switch
strategy based on AIN and the downsized and distributed
form of telephone switching, Cisco and Ascend could target
the multibillion-dollar market that today goes to Nortel or
Lucent almost by default.
Why, die-hards will ask, can't these announcements of SS7
deals be viewed as a sure sign of convergence or the
righteous victory of voice over IP against the heathen
time-division multiplexing voice establishment? Simple:
Because Class 5 switches are used mostly for local calling
and feature support, and those functions don't really offer
any value foothold for the data IP players. A local call is two
analog loops linked at a single switch. Why move to IP to
save transport costs when no transport is being consumed?
Even if convergence isn't what we could delicately call crap
(which it is), it won't influence the outcome of the data
communications company-vs.-telco fight. The majority of
the profits Lucent reported in its recent third-quarter
announcement had nothing to do with any product that
would be affected in any way by convergence.
What would affect Lucent is a change in edge switching, and
that's a market the data players can't easily turn into a data
market. We have mouths, ears and analog phones, not
RS-232 interfaces.
One thing the convergence advocates say is true: We are at a
pivotal point in the evolution of the voice market. The big
voice players have a revenue stream that literally dwarfs data
revenue. If those voice guys get into data, which Nortel and
Lucent clearly intend to do, then today's data players will be
a revenue pimple by comparison - unless they can carry the
battle to the telco giants' home territory - switching, not
transport infrastructure.
Can Cisco and Ascend see past the convergence and
voice-over-IP mumbo jumbo? Will some other player do that
and ace out the kingpin data vendors and big telco players?
Only time will tell.
Nolle is president of CIMI, a technology assessment firm in
Voorhees, N.J. He can be reached at (609) 753-0004 or
tnolle@cimicorp.com.
nwfusion.com |
