Voice over Internet Protocol or VoIP has been one of the most talked about technology changes of the last decade, so what is all the fuss about? Ask the man on the street (and my brother) and he will probably tell you that VoIP saves you money. This can be true but isn't as simple as it sounds:

 

Breaking a voice call into packets requires a header for each packet and so adds overhead, which equals bandwidth, and so for a single call you need more bandwidth for VoIP than TDM. One myth busted: VoIP doesn't use less bandwidth it tends to use more aggressive compression protocols simply because they need to and TDM networks don't. However, there are ways of winning this bandwidth back: People don't tend to both talk at the same time in a single conversation - except in New York - so this bandwidth, nailed up and of no use in TDM (Time Division Multiplexing or digital voice), is available in VoIP. You also don't have to send packets when you are not talking, by using techniques like Voice Activity Detection (VAD), however this is not a totally effective technology as it is imprecise as to when you should stop sending packets and then start again; known as 'shoot/don't shoot syndrome - shout 'don't shoot' and I don't send the first word.......

 

The bigger and more converged the environment, the more effective VoIP becomes. One of the best consultants I have ever worked with, Diane Halliwell, happens to be a traffic expert in her spare time. Her favorite traffic truism is 'The more traffic you throw at it, the more efficient it gets'. So if I can converge all my traffic on to one network it will be more efficient? Absolutely; the first and most aggressive adopters of VoIP were the long distance carriers where they have massive amounts of traffic of all types; so score a win for VoIP there; you have been using VoIP for a lot longer than you knew. Where VoIP also really scores is in the deployment of new home voice services. Vonage and Skype would not have got anywhere if they need to build a network especially in the last mile, instead they use the fact that you are paying someone else for Internet access over that network and you don't mind paying them to ride it - I sense a later blog entry on the business case for carriers forming.

 

But I am an Enterprise Management Consultant so this is of little interest to me professionally. In the Enterprise the value proposition is different; higher value users and lower tolerance for down time. My traditional PBX uses one cable for the phone and a separate one for LAN; avid readers of this blog will leap to the last entry and note that this is use of physical convergence and so is a cost saving. Sorry not really: Single cable = single point of failure. As this is only a relatively small cost saving; many enterprise users of VoIP still use separate cables to the desk-top. I also am taking on an additional cost as I used to have a home run cable and now I have to put LAN electronics on the cable. In addition, TDM PBX's run - and look for that matter - like refrigerators; LANs suffer from far greater reliability problems for a variety of reasons; more on that in later blogs.

 

So an in an enterprise environment, I take a very reliable product and make it less reliable for a greater expense....with little or no feature increase? Actually yes, that's about the size of it; read back, I said that VoIP is one of the most challenging and least rewarding areas of convergence.

 

So why do people do it? There are really two reasons: 1. IP Telephony is much more than just convergence in the transport and 2. you can't buy TDM PBXs any more. Intrigued? Want to know more? Keep logging in friends, I will deliver more gems on IP Telephony in the next entry.