TDM Replacement: Slow and (Hopefully) Steady

With all the potent discussion related to software-defined networking (SDN) and network functions virtualization (NFV), it's easy to lose perspective about what is happening in the legacy side of the telecom world. However, given we consider this is still a cornerstone of the telecom world, Heavy Reading continues to track the replacement of circuit-switched TDM local end offices on a global basis.

Our most recent update of the IP Network Transformation Tracker sadly confirmed the trend of the past two years: that TDM replacement continues to be a slow process.

As a proof point, using North America as an example, currently only about 31 percent of 152 million lines are VoIP or Packet Switched (PS). By 2017 this gradually improves to 42 percent based PS lines, but some of the growth is misleading given the total base declines by approximately 13 million lines. This level of decline is a common trend in many regions.

Source: Heavy Reading IP Network Transformation Tracker July 2013
Source: Heavy Reading IP Network Transformation Tracker July 2013

I understand why factors such as declining fixed and voice revenue and OTT application impacts make fixed-network investment a lower priority, but what concerns me most is the lack of a well defined transition strategy. Is there a global strategy to replace or simply invoke the sunset clause?

While there is some potential that NFV could play a role in providing a lower-cost approach to replace fixed switches, I have yet to see tangible commitment to making this happen. And if NFV is positioned this would likely lengthen the replacement process beyond 2017 given the incubation time NFV requires for commercialization.

Plus, set against this backdrop is the reality that by 2017 some TDM switches will be approaching 40 years old and still providing key services such as 1-800 and Centrex business services. While these switches are still performing on a steady and even keel, the concern that failure rates will increase is real with the potential for prolonged outages due to lack of spare parts, declining vendor support and the rapidly declining pool of in-house technical staff that understands in detail what happens on the peripheral and control side of a TDM switch.

To be clear, I don't believe we are yet approaching the panic stage predicted in E. M. Forster's seminal short story, The Machine Stops, but we need as an industry to develop a technology strategy of what happens to fixed services when the existing machines that support them reach "full stop" retirement age.

— Jim Hodges, Senior Analyst, Heavy Reading

brookseven 8/23/2013 | 4:57:32 PM
Re: A view with a room Humorously, AFC's original IOC business was really started in the early-mid 90s with the replacement of mechanical and stepper switches...particularly in RSUs.

In the old days, keeping up the rate base worked to make the switch.  Now its availability of spares.


jhodgesk1s 8/23/2013 | 12:20:20 PM
Re: A view with a room Dennis, I would say operators in the Netherlands and Japan are the current leaders in terms of VoIP adoption.

Carol Wilson 8/23/2013 | 11:56:53 AM
Re: A view with a room When Bell Atlantic bought GTE back in 2000, a former GTE employee swore to me there was still a step switch working in the backwater of GTE's rural territoy. I never confirmed that but it wouldn't have been all that surprising.

Unless there's a business incentive to divest the TDM gear, it could exist in some areas for a while to come. 
mendyk 8/23/2013 | 11:31:00 AM
Re: A view with a room What about the operators?
jhodgesk1s 8/23/2013 | 10:49:20 AM
Re: A view with a room Steve, the vendor footprints vary from region. For example, Genband acquired the Nortel TDM assets so it is a major VoIP player in North America given Nortel had a large footprint of DMS-100 switches. Metaswitch and Sonus are also still in the mix with VoIP solutions in this market.

NSN was a strong player in Europe given Siemens TDM fixed line switches were heavily deployed, but NSN has moved away from this business.

Huawei is the VoIP leader in Asia.
Net Worthy 8/23/2013 | 10:31:26 AM
Reliable old TDM The old TDM hardware is Really Freakin' Reliable. Combined with declining demand for TDM and there's little incentive to replace it. But what do I know, I haven't worked on TDM networks in nearly 20 years.

What about TDM for cellular backhaul? TDM transport is much more expensive per byte than conventional packet-switched networking. I know it's possible to do clock recovery when carrying TDM over packet networks. Is that how it's being done these days?
Steve Saunders 8/23/2013 | 9:48:34 AM
Re: A view with a room Jim, what a great blog. Who is furthest behind with replacing the heavy iron? Who is in the vanguard?

I rather imagine that the trailing pack squirms a little in briefings when questions about 40 year old telecom equipment comes up!



mendyk 8/23/2013 | 8:55:45 AM
A view with a room E.M. Forster makes his debut on Light Reading -- nicely done. No doubt the incumbents are conflicted on when and how to put TDM to sleep. I don't know that a five- or ten-year plan is tenable. For Verizon (to take one example), it may come down to a tipping point on FiOS uptake, almost on a CO by CO basis. Or an event like Sandy can come along and make the decision not to rebuild the old plant an easy one.
Sign In