There Goes the Cellular Neighborhood
- By William Flanagan
- Did you notice?
- It wasn't a big bang, just a mixed series of product introductions, service announcements, and speeches that appear to point the way for the voice network of the future.
- If you want fast Internet access on fiber to your premises, be prepared to give up powering your phone from the central office battery and settle for 99.9% uptime rather than the "five nines" of the PSTN (Verizon FIOS). The difference is about 8 hours per year of outage.
- Carriers who resisted voice over frame relay recently agreed with equipment makers on a standard for handling voice over IP on 4G (LTE) data networks (Voice over LTE, VoLTE; GSM Association).
- The telephone company that isn't a telecommunications service is working with Verizon to put an always-on VoIP client on cell phones that can use a 3G or 4G data plan for worldwide voice calls (Skype).
- 4G radio service--either LTE, WiMAX, or both--will soon cover the globe with enough capacity for a video feed to every phone at once (thus providing bandwidth for a few billion phone calls, too).
- SIP trunking gains respect as an enterprise service.
- The common thread is that each item depends on Voice over IP. Each item reinforces the others. The trend to "all IP" is, if possible, gaining momentum. Ten years ago I didn't consider an IP network suitable for everything, particularly not voice telephony. Now the standards have developed, network capacity is vastly increased, and a new generation of users values mobility above reliability.
- So here's a take on where we're going.
- The avalanche of traffic moving to IP networks (including MPLS and their supporting Layer 1 optical components) likely will sweep away not only circuit switching for voice but also the POTS land line. Why pay to install and maintain copper loops any more? The upside for speed over copper, despite the leaps in DSL performance, will never exceed the capacity of 4G wireless. What carrier really wants to pay to power those stationary lumps of low functionality called desk phones? Eventually what we've heard for years will be true: There's no real need for a dedicated voice network; voice is just another form of traffic on the IP network. Voice (and real-time video conferencing) will need low latency so the network will have to prioritize these packets, but we know how to do that.
- Why put in any cable, even fiber, if 4G radio links can provide adequate capacity and acceptable availability? Instead, drop in a picocell/WiFi router, with an LTE/WiMax uplink, for phone, fax, Internet, TV, gaming, burglar/fire alarm, etc. The very product showed up while I wrote this edition: http://www.greenpacket.com/dl_devices.html. Some devices will have their own 4G radios/modems.
- Truck roll? Let the customer buy and activate the picocell like a cell phone. Repairs? Bring the device into the phone store to swap for a new one. That will be hard on the unions.
- Legacy Cellular? Yeah, it will hang on, the way circuit switching did. A few people will use it because they like the simple 2-function handsets with the large numbers on the screen. HSPDA will be enough from many machine-to-machine applications.
- E911? GPS in the picocell or the handset tracks locations, stationary or mobile. This service will require additional infrastructure to route calls to the proper answering point.
- Long distance? How the world has changed. Recall that AT&T, when forced into divestiture, had the choice between Long Lines and the Local Exchange Carriers. They chose to stick with LL--oops. The AT&T name survives today only because the LEC that bought it chose to keep the name.
- Costs of cell phone plans today are determined almost completely by the number of minutes of local air time. Distance no longer matters within the US. Skype on a phone does almost the same for international calling. So what's the future business model for long haul transmission? Perhaps it will look like some other forms of transport:
- Early highways in the US started as private tollways. A few still operate that way, but most roads (including toll roads) are owned and maintained by some governmental body.
- Subways in New York City and many other locations started as entrepreneurial enterprises, generating profits for their owners. During hard times, profits stopped; owners walked away or ceded the "lines" to the government.
- Let's make some assumptions about the future: all transport is IP; net neutrality on the Internet requires equal access to all comers; LECs are based almost entirely on IP over wireless local loops. Won't that make the Internet the natural (least expensive) choice for backhaul from "cellular" base stations? Most of the traffic will be "Internet" or data, much more than voice. There might be some extra margin in charging more for prioritized voice/video packets, but the bulk of traffic (data) won't need that level of service.
- Will any private corporation want to bother with LD? Even today, carriers contend that net neutrality will make the Internet unprofitable. If true, carriers could decide to sell their long haul facilities. The buyer of last resort? Some new government or non-profit agency set up to run the public Internet. Carriers could lease back capacity for base station backhaul. The new "public corporation" will restore the concept of "common carrier" to telephone service.
- A few carriers would probably keep a private wide area network, to offer premium services. But it would be a content distribution service, rather than a telephone company.
- Hey you copper thieves! Bring it on! You're making way for the future. Just leave my neighborhood alone until 4G is fully deployed.
![]() |
In Converged
Networking We Have the Experience |