AT&T has asked the Federal Communications Commission for permission to halt sales of frame relay and Asynchronous Transfer Mode services on or after May 31, 2012, with FCC permission, because of declining customer demand.

 For those of you with good memories, ATM was once thought to be the “next generation network” or the “network of the future,” as Integrated Services Digital Network once had a similar position.
 Those of you who have watched industry efforts with subsequent “next generation network” standards including IP Multimedia Subsystem or Rich Communications Suite will understand why some of us are skeptical about any of the top-down “standards” for next generation communications the global telecom industry develops, from time to time. 
It isn’t that standards are unimportant. It’s just that the industry hasn’t been very good at turning proposed standards into workable and useful “real world” frameworks fast enough to be useful.
In fact, most of the marketplace standards were created because critical masses of end users adopted them. That was precisely what happened with the proposed Open Systems Interconnect model, which was supposed to surpass the “legacy” Transmission Control Protocol/Internet Protocol,” which today is just called IP. 
Some network protocol “sunsets” are trickier than others, but all such transitions are easier when the customer base has dwindled, and when preferences for the replacement products are clear. 
AT&T seems to have concluded the cost of supporting, operating and selling the ATM and frame relay networks now are approaching, if not exceeding, potential revenues. 
Someday we will have to sunset the “Public Switched Telephone Network” as well, in favor of networks using IP. With the “sunsetting” of frame relay and ATM, AT&T also will leave behind the days of voice over frame relay and voice over ATM. 
But network modernization and protocol changes are not the big story here. The big story is that the global telecom business has a poor track record when it comes to visualizing and then commercializing any “next generation network.” ATM was supposed to be that next generation network. It clearly failed. The “IP everything” network displaced the OSI framework. 
So far, IMS has been slow to take off, while RCS is in rather early stages. But a betting person might argue that the chances either one is actually widely adopted, and soon, are slim. Worse than that, the odds of marketplace failure are substantial. 
In the past, when voice was the service that mattered, carriers could adopt new generations of switching in a closed environment, which is one reason switch platform transitions have never “failed” in the market. 
These days, all services exist in open and non-captive ecosystems, where there is less a carrier can do “in a closed way.” In fact, most parts of the ecosystem do not even want to deal with any protocol complexity in the network and transport layers. They just use IP and want to assume all the rest of the network layers just work. 
Carriers want RCS and IMS because it makes the task of crafting and delivering services on top of their transmission and access networks easier. But the rest of the ecosystem doesn’t much care, with the obvious exception of suppliers of hardware and software solutions and platforms to carriers.
 And that is why the task of commercializing any next generation network is harder than it used to be. The ATM shutdown simply is a reminder that major failure not only is a possibility, it might be the most-likely outcome.

Edited by Brooke Neuman