The Chicken or the Egg? - Convergence Defined

Introduction

About a month ago, I got a call requesting my participation on a survey about convergence. "Would you share your thoughts?", the lady asked. I thought about that for a moment, and replied that I wanted to talk about divergence. "No", she said, "I'm talking about c-o-n-v-e-r-g-e-n-c-e", spelling out the word. She went on to explain what convergence was all about.

Even though model rocketry is a hobby of mine, I'm no rocket scientist. That's OK, because it doesn't take a rocket scientist to figure out what the caller was talking about. Yes, gentle reader, convergence is all about merging voice and data onto a single IP-based network.

Still, this sounds all too familiar to me. Although saavy voice network engineers are hard to find these days, as I've stated elsewhere: Sometimes you've got to look back, to see forward!

Another 60's Flashback

So, Mr. Peabody, lets go into the time machine and go back to the early 1960s. At this time, AT&T begins to develop and deploy the T1 Channel Bank, about 25 years or so after Alex Reeves first conceived Pulse Code Modulation. A wonderful idea, the use of this quot;new" T1 technology was only available to AT&T and the US Government.

Enter the T-span

Now, let's go back into the time machine, set the date to the end of 1983, and look at the state of the customer network. At this time, 56K DDS supported high-speed data networks. (How ironic that these "high speed" data networks managed to get the job done). Technological advances and price reductions have put PBXs in the hands of medium and large-sized businesses. Phone services to the PBX were routed via separate cable facilities to the local carrier. If a business had really big voice requirements, the carrier would be a big Channel Bank cabinet right on the premises.

1984 enters with a bang, and the divestiture of AT&T's Regional Bell Operating Companies (RBOC). Lest I forget our local carrier here, special provisions were applied to GTE, an independent carrier. Now, T1 services start to become available to the commercial sector. Why now? Why weren't T1 services tariffed back in say, 1963, 1973, or 1980 for that matter? Faced with new long-distance competition, a seed is planted. Merge your private voice network with your data network, the ICs exclaim. A new T1 multiplexer market emerges. PBX networking technology flourishes. Voice and data systems c-o-n-v-e-r-g-e.

The Local Exchange Companies start to deliver local phone service over T1 facilities while the Interexchange Carriers (IC) continue to deploy and upgrade their networks to digital technologies.

The VPN: Diverge, and Change Nothing

OK, back into the time machine. Now, we're going to fast forward a few years to 1989. Since the IC market has settled down, ICs now search for ways to differentiate themselves. Sprint deploys a unique service called a Virtual Private Network (VPN). AT&T counters with their own similar service called Software Defined Network (SDN). In both cases, we can now c-o-n-v-e-r-g-e our private voice network with the public network. An ingenious move, these VPNs increase IC network efficiency dramatically, while the end-user moves voice traffic off their private, leased lines. I call this d-i-v-e-r-g-e-n-c-e, since private voice network traffic no longer traverses the same path as the data traffic. From the customer's perspective, nothing has changed. Long-haul voice and data traffic still leaves their facility on a single T1 line.

The Local Exchange Companies continue to deliver local phone service, but through regulatory efforts, this traffic is carefully segregated from the long-distance and long-haul data traffic.

Fundamental Network Laws

There are three fundamental network laws that apply to customer networks:

Networks never stay the same
Networks never grow smaller
Bandwidth (DEMAND) never decreases

In Carrier networks, two other fundamental laws apply:

He who has the most stuff in the ground wins
If you build it (SUPPLY), they will come

These fundamental laws present an interesting economic model. Now you're catching on...

Conclusion: A Rose by any Other Name

That IP voice is here is a foregone conclusion. But it's easy to get mired down in all the particulars of the Internet Protocol. VoIP, at best, is at a point where all Internet and voice traffic traverse a common IP-based network(s). Of course, let's make sure that we run RSVP protocol (a glorified term for overhead) so we can get decent quality. Whether you negotiate RSVP throughput and SLAs with the IC, or order additional DSOs between your customer facilities, the net result is the same.

Convergence? Pooh. Dear reader, you never really left.