Hybrid Networks

Bayly Communications White Paper
Version 4.0, October 22, 2004
Network convergence over IP – one size does not fit all!

Originally, the compelling cost savings of consolidating voice, video and data
transmission over a single network sparked interest in network convergence over
Internet Protocol (IP). Today, network convergence over IP is largely being
driven by the promised cost savings of IP Telephony. In fact, In-Stat/MDR
predicts that the number of IP lines shipped in PBX systems will grow from 5.5
million in 2003 to 15.9 million in 2008. In the same report, they predict that while
the PBX market is mature and slowing, between 2003 and 2008 shipments of
converged PBX lines will grow at a 11.2% CAGR, while pure IP PBX line
shipments will grow at a 28.9% CAGR.1
This is leading many organizations to create infrastructures based entirely on IP.

But is this the right solution?
Convergence should not necessarily equate to pure IP

Through aggressive marketing by IP vendors, general market perception is that
convergence is synonymous with IP. In a recent article in TMC Internet
Telephony2, author Tony Rybczynski defines convergence as follows: “Network
convergence is the act of bringing voice, data, and video onto an IP, Ethernet, or
optical network. Enterprises striving for uniformity have focused on the IP
protocol suite, as the protocol of choice for networking and applications, spurred
largely by the Internet and by the economics of having fewer protocols to
manage.” This article and others like it suggest that IP is the protocol of choice:
we believe that many I.T. professionals take this to mean that IP is the only
choice. In fact, for certain applications, it’s definitely not the best choice.
We define convergence a bit differently: “Network convergence is the migration

of technologies from legacy based voice and data to IP based technology. This
will be a gradual transition characterized by consolidation and optimization of
existing network infrastructure where users will need to support a mixture of
technologies on a common platform.”
Critical information needs better than best effort

Critical infrastructure networks are those that require guaranteed, always on,
service. They all have very similar requirements: they carry corporate non-
critical traffic that can capitalize on the best path routing of IP — but more
importantly, they have critical traffic requiring guaranteed connection. They
typically have a separate network that carries their critical data over fiber and
microwave – using transport mechanisms such as SONET/SDH (TDM), ATM or
Frame Relay.
1
IP PBXs: A Market Hitting The Tipping Point, May 2004, In-Stat/MDR
2
“2004: The Year of Convergence”, Tony Rybczynski, TMC Internet Telephony, January 2004
Today, IP is known as a best effort network, that is, it does not provide
guaranteed delivery. While this may be more than adequate for most enterprise
communications, for critical infrastructures, the network must be “always on.
Some examples of critical infrastructure networks include:
• Transactional information in financial networks such as banks, insurance

companies, etc. In these environments, a network outage or traffic
slowdown can translate into literally millions of dollars lost.
• Monitoring information in industrial and transportation environments with
programs such as SCADA – real-time data gathering from remote
locations to control equipment and conditions. Imagine the problems if
data packets from a train monitoring systems did not arrive properly – it
could result in track problems, causing accidents and even death.
• Voice communications in enterprise networks – while best effort Voice
over IP (VOIP) may be acceptable for many organizations; voice is very
sensitive to latency, which can cause echoes, or even worse, a delay in
hearing what the other person is saying. In fast-paced business
environments where good communication is essential, this may not be
good enough.
• Monitoring systems in industrial applications such as oil & gas production
and distribution. Reliability for SCADA applications is critical to ensuring
that equipment in the field is operating properly — imagine the
environmental implications if a report of a damaged or blown wellhead
can’t reach the central monitoring site!
We believe these critical infrastructure networks — such as utilities, oil & gas
companies, financial institutions, government and transportation authorities —
need to take a hard look at a hybrid approach to convergence, for several very
compelling reasons:
• They have guaranteed delivery requirements for critical monitoring data
and voice that just can’t be answered with IP.
• They’ve already invested significant funds in their infrastructure, which can
be extended for several years by using a hybrid approach.
• The infrastructure of leased lines and private fiber networks already exists,
and can be effectively leveraged.
• In some of these private networks (like jungles & deserts), the Internet
may not even available, which could negate the benefits of using IP as a
transport mechanism.
• Their critical infrastructure can be vulnerable to Internet-based threats,
which puts the entire organization at risk. For some organizations, this is
not an acceptable risk.
Real Life Rationale for Hybrid Networks
On-time, accurate signaling is key

In railway systems, getting the right signal to the system at the right time is critical –
people’s lives could depend on it. In these networks, security and reliability are key –
can you afford to lose communication as a result of a virus, worm, or other malicious
software that can infiltrate the network through the Internet? Consider this story…
"A computer virus was blamed for bringing down train signaling systems
throughout the East today. 'The virus infected the computer system at CSX Corp.'s
Jacksonville, FL, headquarters, shutting down signaling dispatching and other
systems at about 1:15 a.m. Eastern time,' CSX spokesman Adam Hollingsworth
said."3
Down time unacceptable

For many organizations with critical business operations, down time can prove to be an
expensive issue. Recently, Merrill Lynch decided to protect themselves and ensure
“always on” network availability and reliability. An article in Business Communications
Review reported that “… Merrill Lynch announced it was replacing Cisco VOIP gear
with a hybrid IP-TDM system from Avaya in its Hopewell, NJ location, citing security
concerns that would seem to go to the heart of the convergence issue.”
A Merrill Lynch spokesperson indicated “… there is an increasing concern associated

with the risk of completely losing both voice and data communications as a result of an
IP network outage. This exposure is amplified by the constant presence of IP security
vulnerabilities, and the increase in the level of attacks over the past two years.”4
If IP offers only best effort, then why use it for critical traffic that really requires
guaranteed delivery? It makes more sense to create a converged infrastructure
that utilizes the transport mechanisms that meet specific needs, particularly for
critical applications such as industrial, transportation, utility and finance — much
of the critical infrastructure is already in place! By doing this, organizations can
prolong legacy investments while implementing solutions in a controlled manner
that makes sense for their business.
3
The Associated Press, 8/21/03
http://www.cbsnews.com/stories/2003/08/21/tech/main569418.shtml
http://computercops.biz/article2856.html
4
Business Communications Review, September 2003,
http://www.bcr.com/bcrmag/2003/09/p26.php
Building a hybrid converged network
We believe that organizations should seriously consider building a hybrid
network, versus converging all network traffic onto IP. Existing network
infrastructures should be evaluated, and either phased out or optimized into the
hybrid network, potentially as part of a longer-term network evolution plan.
Any converged network – whether a hybrid of legacy and new technologies, or

exclusively IP – must satisfy several key business requirements:
• It should support legacy voice and data, so that organizations can
leverage these often-expensive investments for as long as possible,
minimizing capital expenditures.
• It must support a multi-vendor environment, to allow organizations to
transport all traffic, and use the solutions that best meet specific needs.
• It must provide the flexibility to support both bandwidth optimization (best
effort) and guaranteed delivery.
• To ensure guaranteed delivery and equipment optimization, it should
support mixed technologies including low latency TDM on SONET/SDH,
legacy voice and data, ATM, Frame Relay, Multiprotocol Label Switching
(MPLS), and of course, IP.
Encourage evolution, not revolution

While prevailing opinion suggests that pure IP is the way to go, we believe that
organizations should take what they have today and migrate deliberately and
cost-effectively, creating a hybrid network.
Before throwing out existing network equipment in favor of an all-IP

infrastructure, organizations should objectively assess their business needs, and
ensure that they match their network requirements to select the appropriate
network elements. We believe the best network convergence solution for critical
infrastructure networks is a hybrid approach: one that leverages the benefits of
IP, but also that complements other legacy network technologies with the goals
of improving network efficiency, consolidating network elements and optimizing
network performance. The result is a hybrid network that provides all the benefits
of IP while maintaining existing infrastructure to provide the right solution from a
business and technology perspective.

Hybrid Networks

Uploaded by

Copyright:

Available Formats

Hybrid Networks

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hybrid Networks

Uploaded by

Copyright:

Available Formats

Bayly Communications White Paper

Version 4.0, October 22, 2004

Network convergence over IP – one size does not fit all!

This is leading many organizations to create infrastructures based entirely on IP.

Convergence should not necessarily equate to pure IP

We define convergence a bit differently: “Network convergence is the migration

Critical information needs better than best effort

Some examples of critical infrastructure networks include:

• Transactional information in financial networks such as banks, insurance

On-time, accurate signaling is key

Down time unacceptable

A Merrill Lynch spokesperson indicated “… there is an increasing concern associated

Any converged network – whether a hybrid of legacy and new technologies, or

Encourage evolution, not revolution

Before throwing out existing network equipment in favor of an all-IP

You might also like