IP telephony: Coming, ready or not

Inevitably, every CIO, IT manager, facilities manager or whoever care-takes the data and voice infrastructure, will spend some time considering IP telephony (IPT).

That time, in some cases, will be as long as it takes to say "It`s too expensive" or "there`s no benefit for us". This may or may not be true, but all too often, decisions are taken without the proper information being provided. I vaguely recall a time when you could sell new technology to many for that reason alone. It was new and "Gee Whiz" and the data processing manager (remember them?) could befuddle the custodian of the purse-strings enough to get some new toys.

Inevitably, reality has come to town and every technology purchase nowadays is preceded by extensive product research, an in-depth business case and a microscopic analysis of the service partner`s offerings. Finally, culminating in a presentation to the board of directors, which has three questions: How much does it cost? How much is it going to save? How much more revenue will it bring us? You have to get at least two out of three of these questions right otherwise it`s the walk-of-shame and forget the new toys.

The process of exploring new technologies, applying the findings to your specific environment, building a strong business case that includes total cost of ownership models is long and arduous. Once this is all done and if you have been successful, the next challenge will be to define an RFP and then evaluate all the responses followed by protracted contract negotiations.

The purpose of this article is to try and assist the ICT executive who knows that IP telephony may hold some value for his company but is not sure what must be done to prepare. The following observations and suggestions are a percolation of research findings from various research organisations.

Get a good working knowledge of what IP telephony is all about. VOIP is obviously a very significant consideration for most organisations as it offers some very attractive cost saving benefits. Some of these benefits are immediately visible, ie toll-bypass, but what about the other benefits such as infrastructure consolidation? What effect would it have if you only had one Ethernet point to every desk carrying both voice and data? How would this play-out from a support cost perspective and from a moves, adds and change viewpoint.

From a productivity angle, VOIP allows automated mobility and collaboration facilities as examples. How valuable would it be if less incoming customer calls ended up in voice-mail? Would online conferencing and collaboration reduce travel costs, save time and generally improve productivity in your environment?

These are some of the sweet-spots of VOIP but what about the negatives? IP is the bloodstream of the Internet and we are all aware of how pervasive viruses are nowadays. Do you want to expose your telephone infrastructure to the same vagaries that your network occasionally suffers from? What are the ways that you can prevent this and what kind of guaranteed uptime can you expect?

If you believe there is an area or areas that can benefit the business or are considerate of the global trends and want to be prepared for IPT, a business case is a good point of departure.

Your current investment in a legacy telephony system if not fully depreciated will have a value that cannot be ignored. What possibility is there for the existing technology to migrate from legacy TDM architecture to VOIP and how cost-effective is this?

Assess your existing data network`s suitability for carrying VOIP. From a LAN perspective, hub-based networks are non-starters as are 10 meg LAN segments, Fast Ethernet is the minimum requirement. From a WAN viewpoint, sufficient bandwidth capacity is obviously essential, quality of service (QoS) becomes very important due to the nature of voice traffic. Latency and jitter can dramatically reduce the quality of your VOIP implementation, therefore a thorough analysis of the WAN traffic is vital. Not only identifying peak times, bottlenecks, etc, but also a more granular assessment of the applications being used across the WAN. What traffic is essential and what is recreational and what can you do to control this.

The objective is to justify the investment in IPT based on cost savings as well as performance enhancements. Quantify these in detail pertinent to your specific scenario. As a simple example, intra-company (between branches) voice and fax traffic information should be gathered and analysed. The potential cost savings here alone could justify the entire VOIP migration. Other areas to consider are reduced head-count due to single-point call attendant, reductions in installation and maintenance costs and overall productivity gains.

Customer service levels can be greatly enhanced with IP telephony features such as find me/follow me, multimedia conferencing and high-power call centre features. These enable agents to provide more focused assistance to customers by providing the agent with customer history the instant the call enters the system.

Unified messaging is another area of consolidation allowing the establishment of a single platform for all message types. These would include e-mail, voice-mail, fax and pager messages in a single unified mailbox.

From a cost consideration perspective, make sure the business case takes into account costs for any additional facilities requirements such as UPS, equipment racks, cabling, etc. It is also vital to include adequate security measures to prevent viruses and the like from entering the network via the proposed VOIP system.

Having established that there is value in VOIP for your organisation, compile an RFP based on your findings. Highlight what features of VOIP are critical in your environment and indicate what base you wish to work from. Thus, if you intend extending the existing infrastructure based on the incumbent vendor`s technology, you can obviously be product-specific. If, on the other hand, you intend working from a zero base, ie, a completely "greenfield" site, specify how this must integrate to the existing network or if you are prepared to replace that as well.

The critical elements of any VOIP installation are:

* Overall capital cost
* Quality of the voice in your scenario (related to your available bandwidth)
* Overall reliability of the service - expect 99.999%
* TCO of the implementation and the corresponding ROI
* Future scalability
* Integration to existing PBXs and facilities for phased migration
* Adherence to current and emerging standards
* Features as required to meet business/productivity goals
* QoS mechanisms

Point two above relating to voice quality is always difficult to assess as the network characteristics vary depending on traffic loads. Additionally, what is acceptable to one, may sound unsatisfactory to another. The ITU-T specifies two standards to determine voice quality. Firstly, the Mean Opinion Score or MOS (ITU-T P.800) is a scale from 0-5 with anything above 4 being considered equivalent to normal PSTN quality. The second standard is Perceptual Evaluation of Speech Quality or PESQ (ITU-T P.862) which specifies a scale from -1 to +4.5, the latter being considered perfect. There are diagnostic appliances and software available to help evaluate voice quality against these standards.

The final stage of the RFP process should involve a pilot-site where you get to evaluate the technology you are intending investing in first-hand. During that pilot, measure the voice quality using a diagnostic tool from Agilent or similar. Assess the management functionality and the ease-of-use, how easy is it to do Moves/Adds/Changes. Test the features in a real world situation, ie set up a multi-party conference and allow certain personnel to utilise mobility features. Most importantly assess the prospective integrators knowledge and willingness to assist you in the pilot. If they are unsure and defer to the vendor continually or their personnel seem reluctant to assist you in exploring the features and performance, you would probably be wise to look elsewhere.

As a final statement on the impact VOIP is going to have, it is interesting to note that the FCC in the US has been approached by most of the carrier operators asking for protection in the form of regulations against VOIP. Unfortunately for the carriers, the FCC views VOIP as "information services" in most instances and can thus offer them little refuge from the coming onslaught.