February 2003

Migrating From Circuit To Packet: The Business Case For IP Telephony

BY CHARLES HENDERSON

Flashback to February 2000. The Dow Jones and NASDAQ stock indexes were at all time highs, as were the stock values of every publicly traded telecom company. Competitive carriers and the looming Internet telephony revolution were driving long distance rates down to single-digit-per-minute rates. The dot com euphoria spread to the staid world of corporate telecom in the form of a magical new technology that would make the PBX obsolete. Voice over IP would create returns on investment so compelling that every corporation would begin replacing their traditional telephony systems. Indeed, standards-based technologies promised to make the need for corporate telephony systems disappear, along with much of their cost.

Telecom manufacturers were pouring millions of dollars into creating Voice over IP (VoIP) systems to stave off the implosion of their installed bases with their lucrative service contracts. They were also pouring money into marketing campaigns designed to convince corporate telecom buyers that their products weren�t PBX�s after all but Internet-ready converged platforms.

Oh, how the telecom world has changed in three short years. The dot com�s are gone; except for the few who had old world (read profit-making) business models. The Dow and NASDAQ are beginning to show life after flatlining for almost two years. Corporate earnings in the telecom sector are beginning to crawl off the bottom. But, the phones on most people�s desks are the same as they were three years ago; so is the technology in the basement that runs them. PBX is no longer an anathema and even the IP-only technology vendors are promoting a migratory approach to convergence.

The state of the industry in 2003 is one of stark reality. There is no magic. And real, achievable return on investment analysis is driving technology acquisitions. For telecom departments this means that the existing infrastructure that is delivering high-quality, highly reliable, very secure communications at predictable costs will continue to serve the needs of most corporate users for its normal lifetime. As new sites are added to the network and as systems are replaced or upgraded, IP-based technologies will be one of the tools considered for the job. And over time, as data networks migrate to full real-time quality of service capabilities the trend will be to implement IP.

The challenge for the telecom manager contemplating the future is when, where, why, and how are IP technologies part of the solution. The answer is, as always, in the why.

VOICE AND DATA NETWORKS ARE BOTH UNIQUE

For a long time the conventional wisdom, from the data networking point of view was that data was hard and voice was easy. Of course, the voice networking point of view was exactly the opposite and both are right. And both are wrong.

The reality is, that despite the methodology for transmitting the payload, the knowledge required to deliver high-quality voice networks is distinct from that required to deliver high-quality data networks. While the mission of the transmission network is similar, the mission of the end-points and the operational environments are unique. Thus the skills required to deliver the functionality required by the end-user does not converge just because the packets are intermingled.

VOICE AND DATA NETWORKS MUST BOTH EVOLVE

Many of us in the telephony world assumed that corporate data networks were well defined, well managed and properly implemented. What we found was a highly dynamic environment with widely differing options for achieving similar ends. It is an environment so dynamic that one implementation was not complete before its architectural replacement was arriving on the loading dock. And the environment is so dynamic that applications, protocols, and even user behavior seemingly changed by the second.

Early assumptions that only the voice components had to change to achieve convergence were naive. The presence of voice in the network places architectural and administrative requirements on it that cannot be compromised or the characteristics of the voice application will not meet expectations.

Understanding the why really boils down to understanding the characteristics that telephone systems have traditionally delivered and what new expectations users have that are changing those characteristics. Telephony users have a common set of expectations of telephone systems based on many years of experience: voice quality, reliability, feature set, lifeline services, user integration, and security.

These characteristics were adequately addressed with the existing PBX technology for many years and were quite stable. But new characteristics have begun to be demanded by consumers including: transport integration, standards adherence, open systems, and integrated applications. These emerging demands from consumers can create a new compelling reason to move to VoIP as long as the real benefits exceed the cost enough to justify the investment.

Five Key Evaluation Criteria

In developing a business case for any large telephony acquisition the first and most important step is to develop a set of criteria that the project must meet in order to proceed. Meeting these criteria will generate cost in the project but also, aside from operational cost savings, they will generate all of the benefits to justify the project. In addition to the financial criteria established by every organization, here are five key areas where requirements must be understood and met by the project in order to be acceptable for implementation.

Constituent Requirements

Your constituents are your customers, and often your customer�s customer. A new system must not only deliver similar functionality to the existing system, but also meet additional needs to create quantifiable benefits that can offset costs. What are the minimum expectations? Are there occasional or seasonal needs to be met? Is telecommunications a mission critical tool of the group? How do constituent groups interact? Can this be improved? Do people need access to service while traveling or working from home?

Reliability and Security

Reliability is now defined as more than just high uptime. Organizations are requiring non-stop telephone service even in the face of catastrophic disasters. Disaster planning in most organizations has taken on significant new meaning in the past year. Packetized voice provides opportunities to provide enhanced disaster tolerance but it also creates challenges to maintaining the reliability and resilience as the networks and organizations evolve. Reliability in the twenty-first century also requires careful coordination between telecom departments and their carriers. Careful consideration and planning needs to take place to identify how calls will be routed in various outage scenarios.

Placing voice on the corporate data network creates openings for eavesdropping, virus attacks, and denial of service attacks that did not exist before. One of the strongest arguments for a migratory approach is the ability to maintain sensitive communications and operations in the circuit-switched arena while the packet capabilities catch up.

Infrastructure Challenges

This involves both the PBX and the data network. Is the data network ready for real-time with Ethernet switches throughout the network and routers only where absolutely necessary? Do all of the network components have uninterruptible power at all locations with adequate battery capacity for the network and all of the phones? Remember, fully 50 percent of a PBX�s power consumption is frequently directed to the phones. Are there other high demand applications that may compete for resources with the telephony application?

Vendor Adaptability

Is the solution flexible to meet your needs or do you have to be flexible to meet the solution�s needs? Does the solution meet your capacity needs today and expected growth for the next 5-10 years? Does the solution have a simple and economic solution for smaller or remote sites? Does the vendor have a strong track record in your industry?

IP Telephony Adaptability

Does the solution provide an IP telephony solution that is equivalent to current circuit-switched features and functionality? Is the IP telephony an economic choice (real ROI)? Will terminal devices be supported for the expected life of the system? Does the solution provide you with a migration path that can mix circuit and packet switched voice? How will this IP telephony solution improve your current service to your constituents?

The State-Of-The-Art In Action

Despite the increased rigor for justification of new telecom projects organizations are finding unique ways to leverage their investments in IP networks to improve service, reduce costs, and protect operations.

A large mid-western state university established a goal of unifying all of its constituents across the metropolitan area where it is located. Initially, this meant replacing three independently operated PBX systems with a single distributed system to serve the main campuses and hospital complex. Due to the high availability and security requirements, particularly at the hospital complex, the campus chose to use available fiber and copper to transport the voice but use a secure IP network to transport all of the control signals. Now that this initial phase is complete, planning is underway to begin rolling out a variety of VoIP technologies including IP telephones to integrate 40 additional locations served by Centrex and key systems, which will use the IP network for both voice and control functions.

A major utility provider on the eastern seaboard had established multiple call centers to prevent a hurricane or other natural disaster from disabling customer service in an emergency. However, coordination between the centers was difficult. By implementing a single distributed call center to serve all of the callers and using the IP network to manage the distribution of calls and network control they were able to increase both disaster tolerance and customer service.

An eastern industrial manufacturing company united all of its locations near its headquarters. By using the existing SONET infrastructure, many voice-only facilities were removed or converted for shared use and significantly greater inter-building capacity was realized. This company plans to begin migrating many of their smaller locations around the world to use IP for both voice and telephones in order to bring greater integration and higher levels of service to its employees. The architecture of the SONET rings provided for new levels of redundancy in the voice network than were possible with the previous implementation.

CONCLUSION

In conclusion, IP telephony offers great opportunities for companies wanting to upgrade and expand their systems. But any plans for change are being driven by real, achievable return on investment analysis in today�s economic climate. To get the best possible system for your organization, you should:

� Clearly understand your constituent needs today and anticipated needs for next 5�10 years.

� Confirm that your telephony solution can meet your capacity, growth, and architecture needs.

� Use an independent third party to perform a network assessment to understand performance issues and possible pitfalls.

� Demand Achievable ROI.

Charles Henderson is Director of Product Management at EADS Telecom. Founded in 1979, EADS Telecom North America has played a pivotal role in every communications revolution�from analog to digital, ISDN, ATM, IP and beyond�designing the systems that helped make these transitions possible. Today, EADS Telecom North America�s Connexity solutions include voice infrastructure, contact center solutions, secure digital radio communications and secure networks.

[ Return To The February 2003 Table Of Contents ]