Networking: Convergence Revving Up

Convergence is a term rendered hackneyed by overuse, much before
it would blossom. Call that growing old without growing up...

As a phenomenon though, convergence continues its march surely
and steadily, promising to remain the underlying theme of all future networks.
The blending of voice and data networks is yet far from over, and the
seamlessness between wired and wireless networks has only partially been
achieved.

Doubt that? Well, look at things from a user's perspective and
you will agree more. As a user, won't you like to receive your mobile as well
as landline calls on a single device? Won't you like to have just one number
and stop bothering about whether you connect to a cellular network or a
fixed-line network, or to both? How about choosing your device depending on
location? Maybe all of that, plus a lot more, eh!

Well, that's probably a big task for the guys who work to make
that happen yesterday for you. Services like call forwarding are probably a
glimpse of such things to come.

Devices like smart phones, PDAs, et al already cater to your
multiple needs like computing, browsing, e-mailing, and telephony. There are
limitations though, largely because the networks are yet to converge
practically.

Yes, convergence may appear to have taken place when we look at
the number of service provider networks. But then it looks grossly off the mark
from an access-side perspective.

So, where lie the gaps?

Well, the gap lies among various standards, networks and
business models; between countries, regulations and interest groups; and most
obviously between what users demand and what vendors and service providers
offer.

Sounds more like chaos than a gap, right? We better discuss
something more addressable, maybe a common challenge.

Mother Internet

Yes, the biggest challenge lies with the mother of all next-generation
networks-the Internet, though that's only natural, as the big opportunities
also lie with the Internet.

Let us look at some of these mega opportunities: VoIP, whose
biggest lure is perhaps low-cost long-distance telephony. Also, MPLS-based IP
VPN, which is fast emerging as a secure and cost-effective alternative to leased
lines for enterprises.

The Internet was not designed to be a telephony network. It was
not developed to be an enterprise-wide secure network either. It was a god of
smaller things, whose ability to adapt to other network environments and
simulate those has been striking. This has drawn a flux of innovations to the
Internet and helped develop a whole new universe around it.

Making VoIP comparable with toll quality is arduous not only
because of the complexities of networking involved, but also due to the
conflicting interests of the stakeholders.

Convergence is not just about establishing protocols between the
PSTN and IP networks, but also about having meaningful dialogues between
old-economy carriers and dot-com companies.

Nobody really owns the Internet and so monopoly is not possible
and no one has the financial capability to do so. So while Yahoo, Skype or
Vonage would love to make VoIP toll-quality, an AT&T, BT or BSNL would like
to delay the phenomenon till they are ready to reap its advantages.

Indeed, the Internet is marred by all banes of democracy. But
then it also carries the vital force of democracy with it which keeps the ball
rolling. The Internet economy keeps getting bigger and bigger and so does the
lure, with each passing day.

The New Building Blocks

The digitization of voice and content has gone hand in hand with the spread
of the Internet, both complementing and supplementing each other's growth.
Take for instance, VoIP, for which softswitches and media gateways are some key
building blocks. The adoption of these equipments in service provider networks
has been somewhat slow so far, but is expected to gear up in the near future.

A big push in this regard came from the $1 bn deal with Ericsson
in August 2006. The deal envisages Ericsson to upgrade Bharti's network with a
mobile softswitch solution that would pave the way to an all-IP network.

Other service providers, including Reliance and Tata
Teleservices, have also deployed softswitches in their networks, though the
scales are different.

Likewise, multi protocol label switching (MPLS) is a key
building block of IP VPN networks. Much has been written about the beauty of
MPLS, which can converse with both legacy and new networks. Its simplicity
allows incumbents to protect their investments in ATM and frame relay networks
while offering IP VPN services competitively.

Beyond Switching and Routing

While switches and routers remain the foundation on which networks are
built, the market is maturing and looking beyond. Also, switches are no longer
confined to layers 2 and 3, but have entered the realm of layers 4 through 7.

Growth in L4-7 switches has been much higher than the overall
growth in switches for the second quarter of 2006 over the prior-year period,
albeit on a smaller base.

Vendors are also realizing that there is more to next-generation
networks than just TCP and IP. Some smaller companies like Netscaler, Redline,
Peribit, Packeteer, etc. have successfully demonstrated how special-purpose
network appliances can take the load off traditional switches and routers and
significantly speed up the network.

These specialized network appliances are designed to do a small
piece of the work. Typically, these appliances perform tasks like WAN
optimization, acceleration, secure content and application delivery, integrated
network message routing, and so on.

Datacenter Boom

There has been a phenomenal rise of social networking sites like Myspace and
Orkut, and the online encyclopedia Wikipedia has been a huge success. The surge
in Youtube's popularity is too well known to discuss again.

In the not-too-long run, this will dictate a shift in the way
the network is developed and aligned. This will lead to a paradigm shift in
networking, though the process itself may remain transparent to users.

More immediately, however, the datacenter switch market will
witness a strong growth, as the Googles and Yahoos of the world are making huge
investments in these areas.

In August 2006, for instance, Google reportedly said that it
would invest at least $1 bn for a 1 mn sq ft datacenter in a special economic
zone in Andhra Pradesh.

Talking of other areas, investments in mobile networks will
continue unabated. The growth story here is too big to be overlooked. BSNL's
$6 bn GSM project, which went to Ericsson, Nokia and Alcatel-ITI combine, has
easily dwarfed all other networking investments so far.

Broadband and triple-play services will also be vying for
limelight in the next year.

Some disappointment may be witnessed on the 3G front though, as
non-availability of spectrum may delay a takeoff.

It will be interesting to see the interplay of these services
and how they contribute to the rev-up of convergence, which is nothing but a
multiplay of services.

Deepak Kumar


maildqindia@cybermedia.co.in

The author is senior manager, Communications Research, IDC India