0% found this document useful (0 votes)
65 views7 pages

Convergence Challenges For Enterprise Networks

A Gartner Whitepaper

Uploaded by

Kunal
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
65 views7 pages

Convergence Challenges For Enterprise Networks

A Gartner Whitepaper

Uploaded by

Kunal
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

White Paper

Convergence Challenges for


Enterprise Networks

May 2001

Gartner Consulting 5000 Falls of Neuse Road, Suite 304, Raleigh, NC 27609
+1-919-850-0611
White Paper
Convergence Challenges for Enterprise Networks

Introduction
Talk of converging voice and data networks is everywhere, but the idea of convergence is certainly
not new. Since the late 1980’s, vendors have tried to integrate telephony systems with data applica-
tions through various computer-telephony integration solutions. Today, the latest Internet Protocol
(IP) centric solutions focus on the convergence of voice and data networks onto a single infrastruc-
ture with significant bandwidth and a single set of management and administrative tools. These
applications will be multi-media, including not only voice and data but also video.

One thing now appears certain; we are all on the path to convergence. Where on the path is not
always clear, but even a cursory look at the research and development priorities of the major players
in this market makes it clear that the next generation of enterprise solutions will be based on a
converged architecture. The question is not if, but when.

Is it time to pull it all together? Should infrastructure convergence precede application convergence
or can they proceed simultaneously? Is cost saving through infrastructure convergence, by itself, a
compelling reason to deploy new network technology? Would the integrated infrastructure make
applications more valuable? Do e-business applications require an integrated infrastructure? These
are critical and complex questions facing businesses today, from small, single location companies
to major, multinational enterprises, and the answers are likely to be very situation dependent. In this
white paper we will describe the key issues and opportunities for enterprises as they navigate the
next few years, and provide some insight into the process of charting the course to a converged
future.

Benefits
So why are we all headed down this path? What are the benefits? Why now?

The primary benefits of a convergence strategy are the cost and operational savings from using one
large network versus two smaller (voice & data) networks; and the ability to easily deploy applica-
tions incorporating voice and data functionality.

Convergence onto a single local area network (LAN) infrastructure has a variety of potential ben-
efits. It can clearly eliminate the need for voice line cards and remote line equipment, and, with
most IP phones sharing the PC’s Ethernet connection, separate, duplicated voice cabling to the
desktops can also be eliminated. Cabling savings alone could be worth $500 per station in some
campus environments. Generally speaking, these savings are only applicable to greenfield deploy-
ments, where the voice infrastructure is not yet in place. There are, however, significant associated
operational savings which will be realistic even in an upgrade or replacement situation. Moves
Adds and Changes (MAC), for example, can be dramatically reduced with the capability for IP
phones to automatically re-register when moved. This could result in savings between $10 and $40
per move for the typical large enterprise.

Entire contents © 2001 Gartner Group, Inc.


Gartner Consulting May, 2001—Page 2
White Paper
Convergence Challenges for Enterprise Networks

In the wide area network (WAN) where most of the leading edge deployment has taken place over
the last few years, savings are largely driven by the elimination of long distance (LD) charges. As
the per minute charges for domestic and many international destinations have declined, the motiva-
tion for this type of deployment has similarly declined. However, even at 2 to 3 cents per minute,
enterprises are still spending over $65B per year in the US for LD voice. In an era when new
“Ethernet Service Providers” are offering distance insensitive, nationwide 100Mb/s WAN service
for $1000 per month, the ability to have internal enterprise voice traffic ride the corporate data
WAN virtually free of incremental cost is still clearly an attractive proposition. In the WAN, as in
the LAN, operational savings are also an important factor. A single WAN solution and single
service provider, for example, results in less administrative overhead and network management
activity for the enterprise.

The second and most important area of benefit from network convergence is in enabling new
applications. New applications can drive not only cost reductions, but also new revenues, and the
value to enterprises can be enormous. While the hype around e-business is subsiding; in its wake, it
is leaving many innovative CRM (Customer Relationship Management) and unified messaging
applications, which blend voice and data to create new value for the enterprise and their customers.
While the computer telephony industry has already found ways to implement some of these
blended applications, the convergence of signaling and carriage for voice and data onto a common
(IP) base makes these applications much easier to develop and implement. For example, the his-
torical methods for networking call centers using dedicated T1 circuits or specialized network
services become much more complex when a CRM application needs to deal not just with voice,
but also customer databases, email response and web interactions. The convergence of all these
media onto a single IP LAN/WAN removes much of the complexity and allows the design of
solutions that are easier and cheaper to deploy and operate. Another example is the emergence of
IP based voice response, speech recognition and voice storage capabilities, which allow functions
such as Unified Messaging, Interactive Voice Response and voice-based information access to be
centralized in a single data center serving a geographically dispersed, large user base. Previously,
these capabilities were tied to a single location by expensive PRI (ISDN Primary Rate) circuits and
had to make business sense purely based on this limited number of users. This flexibility in deploy-
ment dramatically changes the economics and opens up new application possibilities.

Enterprise Challenges
The benefits of convergence, as with all good things, come at a price. It is imperative that enter-
prises move down this road with a clear view of their goals. Indeed, knowing what are the drivers
for convergence is perhaps the most important prerequisite to a sound convergence strategy. Lower
equipment costs may be attractive; but if the installed network equipment already meets the enter-
prise users’ needs, replacing it is an expense, not a saving. Operational savings will eventually be
realized; but in the short term, the transition to new equipment, processes and procedures will
require additional training and implementation expenses. Enabling e-business applications can
drive revenues and efficiencies; but is a converged network truly a prerequisite, and does the

Entire contents © 2001 Gartner Group, Inc.


Gartner Consulting May, 2001—Page 3
White Paper
Convergence Challenges for Enterprise Networks

business case make financial sense? Avoiding further investment in traditional (TDM) PBX or key
(KTS) equipment could also be a motivator; but expansion of an existing TDM system may still
have a better return on investment (ROI) than replacing it.

Despite these cautions, for many companies one or more of the drivers to convergence are already
in place. The time for action is here, and a deployment strategy is required. These companies need
to move from clarifying the goals, to eliminating the potential hurdles. Careful attention to the
selection of vendors, products and deployment schedules can help avoid these hurdles, and the
following should be the key considerations:
♦ Transparency for the end user is highly desirable, unless the goal is to deliberately change their
environment; providing dramatically new features and capabilities. Most leading solutions
claim to have implemented the commonly used features, and it would seem that transparency is
a non-issue. However, the availability of full feature functionality should not be under-valued.
Taking features away from users will guarantee trouble tickets; meeting even 95% of every
users current requirements means 0% satisfied users. Understanding the specific needs of the
users and evaluating the proposed system capabilities against them is critical to avoid signifi-
cant support and re-training costs. In the world of IP voice, particular attention should be given
to support of existing terminals, fax machines and conferencing services.
♦ Availability (up time) of the overall solution should be evaluated carefully. Not only are the
specifications and architecture of the vendor’s products important, but also the inherent avail-
ability of the IP network they use. A single network is good, but it is also a single point of
failure. The requirements of voice are more stringent than most data applications. While the
public network goal of “five nines” 99.999% availability may not be required, “four nines”
(99.99%) availability is a reasonable target for many enterprises, and a challenging goal for
many LAN/WAN implementations.
♦ Data network management takes on new importance, given that the data network is becoming
the infrastructure for all of the enterprise’s critical communications. Competent staff backed up
with solid network management systems capable of monitoring the critical multi-media net-
work parameters (latency and jitter for example) will be core requirements in a converged
environment. Strong vendor support is also important here, complemented with remote diag-
nostic and management capabilities.
♦ Quality of Service (QoS) has historically been a hot issue for voice over IP solutions, and while
it is still a critical requirement, it is not as difficult an issue to resolve as it was in the original
VoIP applications using 56kb/s connections and the public Internet. Nevertheless, even in a
well managed IP network with adequate bandwidth and good latency characteristics, it is
necessary to implement a QoS strategy to prioritize traffic and ensure that voice, data and video
requirements can be consistently met. This may require the replacement of some legacy data
network elements with newer elements that implement the appropriate QoS standards
(DiffServ, RSVP, 802.1p, 802.1q, MPLS, Policy management etc.).
♦ Security needs to be addressed in the same way that QoS is addressed. In particular, the use of
public WAN IP services will probably require the implementation of VPN based security
services capable of protecting the integrity of voice, video and data traffic.
♦ Terminal equipment can also create interesting challenges for deployment. IP Phones are now
Entire contents © 2001 Gartner Group, Inc.
Gartner Consulting May, 2001—Page 4
White Paper
Convergence Challenges for Enterprise Networks

available and competitively priced with advanced sets, but not viable in place of the very basic
sets in common areas. This can add cost for the deployment of gateways and traditional voice
cabling in these areas. In particular, fax machine support can be troublesome, as all implemen-
tations require a gateway adapter, and some implementations simply do not reliably support fax
due to its unique requirements for latency.
♦ Interworking, with the public switched telephone network (PSTN), and with the rest of the
enterprise network is another potential source of headaches. Most systems will interwork with
the PSTN using PRI, but the diversity and proprietary nature of many inter-PBX links may
make the networking of a new voice system with the existing systems a real challenge. The
simultaneous replacement of an entire network of PBX’s is clearly unlikely to be desirable or
achievable, and a hybrid networking strategy will be needed.
♦ New applications, as mentioned earlier, are likely the most important factor in a convergence
strategy. These applications may not be easy to identify, as they are not necessarily being driven
by the IT or Telecom organization in charge of convergence. Typically, however, they will either
fall into the areas of collaboration and messaging where productivity will be the key driver, or
into the CRM/Contact Center domain where revenues, customer service and expense manage-
ment are all potential factors.
♦ Finally, careful attention should be given to the level of support available for these new tech-
nologies from vendors and their channel partners. Convergent solutions clearly demand broader
experience from support organizations. A data oriented support group with no experience in
sizing voicemail storage needs would not be a good choice for installing a unified messaging
system, and a voice installer who cannot terminate an RJ45 jack will not be of much value
installing IP Phones. Companies require vendors and re-sellers with experienced resources in
both voice and data, and preferably in converged network implementations.

Vendor Approaches
After initially leaving this market to the smaller, CTI based vendors, the major players in enterprise
networking (Avaya, Cisco Systems, Mitel, Nortel Networks, Siemens) have made their presence
felt over the last few years with a portfolio of solutions. There are two basic approaches proposed
by the vendors, each with their own strengths and weaknesses. Some vendors specialize in one,
while the established vendors (Avaya, Nortel) offer both options.

Approach 1: Pure IP PBX


Whether based on PC systems for small businesses, or large, multi server clusters for medium and
large enterprises, this strategy implies either the replacement of an existing system, or a greenfield
implementation. These systems have often been designed as IP based solutions from the ground up,
and take full advantage of the cost savings possible through reuse of the IP infrastructure. Incum-
bent PBX vendors offer solutions of this type, for example Avaya’s IP600 (one of their Enterprise
CLass IP Solutions – ECLIPS) and Nortel’s Business Communications Manager. These established
vendors are joined by many newer players, such as Cisco Systems with their CallManager product
built on their AVVID architecture. In this approach, gateways provide interworking with existing

Entire contents © 2001 Gartner Group, Inc.


Gartner Consulting May, 2001—Page 5
White Paper
Convergence Challenges for Enterprise Networks

systems and the PSTN, and the core call processing is generally implemented on standard server
hardware and operating systems. The strength of this strategy is generally a low first cost, and its
primary weakness is the completeness of the telephony feature set compared to established PBX
and KTS systems.

Approach 2: Migration of Existing PBX/KTS to IP


This strategy is the domain of the established PBX vendors. Avaya’s ECLIPS portfolio includes
DEFINITY® IP capabilities to allow upgrade of in-place DEFINITY PBX systems to use IP net-
working, applications and IP phones. Similarly, Nortel’s Meridian® 1 forms the core of their M6550
IP PBX. This is a cost effective approach to convergence, where the core feature functionality of
the current system is preserved, and IP interfaces are added to the system(s) in place to enable the
exploitation of IP LANs and WANs for connectivity. In an upgrade scenario, assuming the existing
systems meet the users’ requirements, this strategy provides the lowest incremental cost and lever-
ages the existing mission critical applications with minimal disruption. In greenfield situations it
can also be used effectively where new locations are to be added to an existing network. In this
case remote IP phones and applications, for example, can be tied back to an existing IP enabled
PBX for a cost-effective implementation.

The strength of this approach is in upgrade scenarios where it is clearly the most cost-effective
solution. The only caution is that enterprises need to be assured of the vendor’s migration strategy
to support open standards and services.

Enterprise Options
How then should the enterprise proceed? Whether they are small, single location businesses, or
major multinational enterprises, they are in need of a business case addressing the following areas:
♦ Clarifying the goals for convergence
♦ Suitability of existing data network to carry voice - associated upgrade strategy
♦ Evolving user needs – what, when, where
§ Features, applications, QoS, security, network management
♦ Feature functionality gaps in current and potential new systems
♦ Strategy for re-use vs. replacement of current network elements
♦ Projected costs for terminals, equipment and support
♦ Financial analysis of the implementation alternatives.
Depending on the results of this analysis, one of three general approaches will likely be adopted:
Deferral, network wide deployment or niche application.

Entire contents © 2001 Gartner Group, Inc.


Gartner Consulting May, 2001—Page 6
White Paper
Convergence Challenges for Enterprise Networks

First, in the unlikely case that the benefits of convergence cannot be identified, or if the evaluation
of the enterprise data network uncovers major issues with no budget to fix them, the enterprise will
have no choice but to defer convergence.

Second, where the benefits of convergence are evident and wide ranging, a network wide conver-
gence strategy should be developed. Depending on the situation, this may include the planning and
budgeting of data network upgrades to meet QoS, availability, security and network management
needs. The current PBXs or KTS may either be upgraded to use this IP infrastructure, or they may
be replaced with new server based, pure IP PBXs.

Third, and most common, specific applications will be found where the benefits of convergence are
clear and justifiable, but an entire infrastructure upgrade to IP is not required. This will result in
niche deployment approach, and an incremental convergence plan without an explicit intent to
migrate the entire enterprise. This implies a slower migration from today’s solutions that is likely to
rely on tactical upgrades to existing equipment.

Conclusions
Convergence strategies can clearly be complex and are highly dependent on specific enterprise
situations. Convergence, however, should not be ignored, and enterprises of all sizes need to con-
sider their options and course of action. The starting point is consensus around the goals for conver-
gence. This provides a framework for the analysis of options and will define the benefit side of the
business case. With the goals clear, the planning team will need to understand their users’ evolving
needs and the capabilities of the current systems to support them. Working with suppliers who have
strong resources committed to the enterprise market, the alternatives must then be laid out and
evaluated against these needs and the enterprise’s ultimate goals. Also critical is to evaluate the
enterprise data network and identify any needed upgrades to meet QoS, availability, network man-
agement and security requirements. The results can then be combined into a business case and
implementation plan.

In Gartner’s opinion, most enterprises should be developing convergence plans over the next 12
months. Where the enterprise’s primary PBX/KTS vendor has a solid IP migration strategy for the
embedded systems, this will usually be the most cost effective strategy around which to build an
implementation plan. In greenfield deployments, or where embedded systems are not capable of
meeting future needs, pure IP solutions should be favored, ensuring first that they meet QoS and
availability requirements, meet the users’ evolving needs, and are capable of inter-working with
existing enterprise systems.

Convergence is here. The benefits are real. Enterprises should pick strategic partners who under-
stand the broad scope of their needs and are committed to meeting them, and take their first steps
toward a converged future.

Entire contents © 2001 Gartner Group, Inc.


Gartner Consulting May, 2001—Page 7

You might also like