T.A. Wallace & Associates, Inc.

Telecommunications & Networking Consultants

 



Member: Society of
 Telecommunications Consultants

The Beautiful Myth [Voice is just another application on the network]
 

Business Communications Review

01/01/2007

Copyright (c) 2007 Bell & Howell Information and Learning Company. All rights reserved.

By: Whitehead, Richard

 

There's a myth-albeit a beautiful myth-that has become firmly rooted as truth in many organizations today. Simply put, IT and network managers who are considering implementing IP telephony systems have been lulled into the comforting idea that voice is just another application on the network. While this sounds logical and would certainly make the transition to a cost-effective converged data model easier, it's simply not true.

I'm not just talking about the analog peculiarities of voice that make it different, although they certainly complicate things. It's the fact that even a simple telephony implementation will probably be the most complex distributed hardware/software application deployed on the network to date.


I have been stunned at how many times I've seen companies assume they can "plug and play" voice, only to be disillusioned by the complexity of it. Even today, most large enterprise applications adhere to basic client-server architecture, and it's tempting to consider an IP-telephony solution in a similar manner. After all, many distributed phone endpoints being controlled by a central server (the IP-PBX) certainly seems to fit.

 

But the signaling between the phones and IPPBX is only part of the story. The phones themselves communicate between each other as peers, they also can connect to multiple gateways and media termination points. It's a communication mesh far more complex than most enterprise applications.

Recognizing the beautiful myth doesn't make IP-telephony adoption more complex or daunting. On the contrary, I believe that understanding exactly what you're dealing with and how it's different from most other standard applications on the network can, in the long run, save time, money and innumerable headaches.

One critical element that makes IP-telephony different from other data applications is the extent to which telephone calls are controlled much higher up in the OSI 7-layer model, with the lion's share being done at the application layer. In contrast, most data-oriented applications rely predominantly on the network layer (OSI Layer 3) to provide authentication and access to trusted elements, while the application itself provides a relatively thinner layer of authentication and access.

 

A typical IP-PBX solution, for example, handles the dialing plan, class of service and other calling permissions as sub-layers of access logic within the application layer. Even though these sub-layers directly affect connectivity, they can't be seen by traditional network management tools, because they exist above the layer these tools observe.

 

Another major IP-PBX application sub-layer with unforeseen effects on network management has to do with call routing. While data applications are typically routed at layer 2 or 3 of the OSI model, VOIP calls can take many routes and depend on business decisions (and associated application-level routing) to determine the actual path taken. This is especially relevant as telephone calls can traverse the PSTN as well.

At the session layer (OSI Layer 5), most decent IP-PBXs implement some form of call admission control to deny calls that would otherwise cause congestion and impair other calls already in the network. Call admission control isn't necessary for most data applications, which are very tolerant of bandwidth constraints and network congestion. At the application level, when the IP-PBX determines that the preferred network path is becoming congested, it can first try to downgrade the codecbeing used, to make the calls more efficient, or it can start offloading them to a secondary circuit, or even block the call completely, and treat it with a busy signal.

 

Conclusion

My objective in busting the myth that voice is just another application on the network is absolutely not to dissuade anyone from considering deploying IP-telephony today. IP-telephony can be beautiful-a vital business solution, cutting costs and allowing for voice and data convergence and contained communications management. It's quickly becoming the standard that all organizations, eventually, will adopt.

But just as the first television programs were not simply "radio with pictures," neither is voice just another application that includes telephony. It's much more complex-and powerful-than that. Only when we truly understand the nature and nuances of IP voice can it fulfill its promise and vast potential

 

T.A.Wallace & Associates, Inc. is an independent consulting firm. We do not have financial relationships with any vendor or conflicts of interest that affect our recommendations. Our consultants and Project Managers have over 150 years of combined experience in the industry. If you want an independent consultant who will work strictly in the best interest of your business, Call us on 508-758-9217 or visit us at http://tawallaceassociates.com