Voice and data are like water and oil - not easy to mix. Nevertheless that is exactly what the IP telephony is all about. Network managers want to cut costs by combining the two disparate infrastructures to carry both voice and data traffic, using IP based transport and protocols.
This raises at least these two questions in my mind:
1. Who really needs this technology and where?; and2. Can the technical challenges be overcome?
But first things first. Here's a summary of my concerns:
· The attractively priced deals offered by the traditional voice carriers to the larger customers make it very unattractive for them to merge the voice and data in a company's backbones.
· The potential savings in operational expenses resulting from creation of a single infrastructure are likely to be eaten up by additional costs associated with the increased complexity of managing it.
· The traditional telephony has benefited from decades of development, offering stability and a large range of sophisticated features. Risksassociated with abandoning this technology in favour of the immature and currently under-featured IP Telephony far outweigh the dubious short term savings.
· The same applies to the corporate LANs. By throwing bandwidth at the problem spots, improving the LAN design and up-skilling the support staff, managers have found the fine balance of running the network without putting out fires on an hourly basis. Unfortunately for them bandwidth does not equal control. Adding voice to the LAN will inevitably result in chaos.
· What about the carriers - do they need IP telephony? The answer depends on the type of technology deployed in a carrier's backbone.
Those who have ATM backbones are likely to use ATM native features to combine the carriage of voice and data. Those with IP based backbones - bingo! This is where they can leverage their data networks to carry voice and reap the financial benefits, assuming the technology will allow them to effectively compete with toll quality traditional voice. Even the stingiest of budgets get increased, when poor quality [of voice] affects the bottom line.
It is worth noting that there are other niche segments that can benefit from IP telephony, if and when it becomes technically viable, which leads to the second question of technical challenges.
At the risk of being booed by my colleagues, I'll say that the technology challenges will be the easiest to conquer, as long as the marketers smell the dollars.
In fact even today Fast Ethernet and Gigabit Ethernet combined with Layer 4 switching in the LAN and WAN already provides an intelligent IP transport that can identify and appropriately carry these disparate types of traffic.
Further efforts by IEEE's Differentiated Services working group will help define enhanced methods of providing different Classes of Service for different types of traffic.
It is only a matter of time and money.
David Gabo is Cabletron Systems' director of technology, based in Frenchs Forest, NSW.