The last time I dedicated this column to a look at the softswitch, it was
June 2000. In the two years+ since then, a lot has happened, so I felt it
was an opportune time to revisit the issue.
That particular article generated a lot of interest and discussion from
our readers. Back in 2000 I thought the term ï¿½softswitchï¿½ was a bit overused
and confused in the VoIP market segment. Over the last few months I have
found it used somewhat less frequently -- most likely because of the
slowdown in the entire industry. Weï¿½ll consider the ï¿½confusedï¿½ part a little
It is nice to see that softswitches are still forecast to be a significant
revenue stream in the near future. Frost &
Sullivan (ï¿½2002 World VoP Equipment Marketï¿½) forecasts softswitch
revenues to grow from $409 million in 2003 to $3,106 million in 2006. And
although I believe the industry has about the same chance of reaching those
2006 numbers as my beloved NY Mets do of winning the World Series that year,
thereï¿½s no question that softswitches are in for significant growth.
STILL A MYSTERY
Is the softswitch still as mysterious now as it was back in 2000? The other
day, I examined a number of vendorsï¿½ softswitch products via a quick Web
site scan to see how they define themselves. Here are some of their
I would claim -- and I hope you agree -- that the softswitch definition is
still fuzzy, broad, and mysterious. In many instances, it depends upon where
a company plays in the network space and where it believes it can make money
as networks continue to merge. Based upon this, I will stick by my original
definition. A softswitch:
- Integrated media gateway, applications, and Class 5 functions;
- Class 5 softswitch;
- Access gateway, Class 4, Class 5, media gateway;
- Agents for signaling/control and servers for OAM and P/billing;
- Call control, VoIP features, signaling interworking, and media gateway
In reality, it seems that the softswitch vendors have accomplished Class 4
toll replacement, while Class 5 is still the end game. Recently, some
vendors have released products that support the top revenue generating
and/or most popular Class 5 switch features. Even so, there is still work
ahead to support the thousands of Class 5 features in the new modular
networks. The industry -- in particular, IP telephony -- would benefit from
a clear definition with clearly defined softswitch functionality.
- Integrates the Internet telephony and circuit-switched worlds;
- Duplicates Class 4 and 5 switch capabilities;
- Operates in the classic public network environment where call control is
separate from media flow.
Why is the softswitch so significant? Letï¿½s take a qualitative approach to
this business case to see if it makes sense. For a network designer, meeting
challenges will require a number of things including optimizing the network,
lowering total cost of ownership, and supporting value-added features. We
will not solve these issues here, but letï¿½s use them to solve the business
In terms of optimizing the network, letï¿½s assume itï¿½s a legacy,
circuit-switched network, with proprietary systems that bundle all telecom
functions into a big, expensive piece of equipment.
The softswitch concept was built on the idea of lowering costs by separating
various functions (call control, switching, and media) to lower the total
cost of ownership. Every piece of equipment does not need to be co-located.
To survive, more value-added features are supported to bring in more
revenue. For example, one of my new favorite applications, becoming popular
in the Asia/Pacific region, is SMS over fixed-line phones. I do not find
this capability significant for an all-wireline phone network, but I do like
the simplicity of merging this with wireless networks. Then a wireline phone
could send a short message to a wireless phone. Since SMS traffic is already
clogging the SS7 network, SMS offload over IP may drive significant IP
investment in the foreseeable future. Applications such as this will drive
capital expenditure budgets toward IP.
I believe the softswitch is the way to economically solve these issues. We
will need a flexible architecture to build out pockets of new
infrastructure, but we may want to consolidate the media functionality in a
separate geography. We will not have to expand the circuit-switched network
to accommodate this, but gateway functionality will be required. However,
the network designer will have a hard time shopping for the necessary
equipment based upon the marketing claims above.
One other point is worth mentioning. Open, standards-based building blocks
have expanded over the last couple of years in both number and
functionality. Densities have increased. And SIP has come into its own as a
key protocol for media control (although issues still exist). I can now get
672 VoIP channels on a single board with echo cancellation and various
coders. High-density circuit-switched boards are available as well.
High-availability and NEBS-compliant chassis almost complete the picture.
These building blocks will better enable integrators and independent
software vendors to cost-effectively build softswitches.
The main issue going forward will be defining the softswitch. If the
industry can really agree on its functionality, then we can move on to more
You can check out for yourself the significance and the mystery of the
softswitch at the upcoming Internet
Telephony Conference and EXPO at the Hotel Inter-Continental in Miami on
February 5-7, 2003.
Jim Machi is director, Product Management for the Network Processing
Division of the Intel Communications Group. Intel, the worldï¿½s largest
chipmaker, is also a leading manufacturer of computer, networking, and
communications products. For more information, visit
To The January 2003 Table Of Contents ]