The scramble to rescue voice-over-IP from the technological recycling bin

If you threw all your black phones in the trash two years ago when the voice-over-IP and convergence hype was launched, you’ve probably long since dug them out and continued with traditional telephony. Vendors are now touting voice over DSL, or “packet voice,” as if IP voice were a dirty word. Has voice-over-IP really been rejected, or are we just learning that some of the things we expected from it were unrealistic?

We in North America are accustomed to very good local-call quality, and long-distance quality that’s nearly as good. With the steady drop in voice charges for traditional non-IP voice services, voice-over-IP’s claim of significant cost benefits becomes less credible. How far can rates fall from a nickel per minute? And with the incumbent local exchange carriers’ decision to deploy ATM in their access/ DSL networks, voice-over-DSL really means voice-over-ATM, so the buildout of broadband access seems to be passing voice-over-IP by as well.

This isn’t a failure of the technology but of our expectations. There was never a real chance that the public voice network would be redeployed using IP. Because that essentially dumb expectation was so publicized, we didn’t glom onto some of the areas where real voice-over-IP potential exists. They won’t justify the market hype, but they might save some vendors and venture capitalists.

One potential area is the cellular market. No matter what people say on commercials, mobile phone voice isn’t equal to wireline voice. Because the quality expectations of consumers are lower, cellular providers have more latitude to deploy IP voice. In addition, buildout of cellular networking is still ongoing, which means we wouldn’t have to displace traditional equipment to adopt voice-over-IP – at least for new carriers.

Public interest in the “wireless Web” may also validate IP in the cellular market. If a significant number of wireless instruments end up with IP browsing and e-mail capability, the network that supports them would be more likely to be based on IP.

Stationary IP voice might be even better, and there’s a play there, too. About 800,000 sites in the U.S. are candidates for an IP PBX or key telephone system in the next five years. Unfortunately, a key to success in this market is the ability to support standard phones and PCs as instruments for making calls. Vendors have all too often added all kinds of new IP gizmos to their offerings in the hopes of raising profits, but all they’re doing is lowering their prospects for selling anything. We need to retune the IP PBX arena to accommodate market reality.

Even voice-over-DSL – or let’s face it, voice-over-ATM – may offer some hope for the voice-over-IP camp. We’re not going to be doing voice-over-IP-over-ATM, but we certainly could use the gateway controller products designed for voice-over-IP to support ATM-based voice calling. The media gateways, convergence switches, or whatever you want to call them are often IP-specific and thus unlikely to be used in wireline voice in North America. The gateway controllers that use IP not to transport voice but to connect controllers to the media gateways are still a play. The IP-centric Internet Engineering Task Force’s Media Gateway Control Protocol is now morphing into Megaco/H.248, an international standard that specifically envelops ATM voice networks in the gateway controller feature set.

There’s even some hope for voice-over-IP in the toll network, in which regulatory and technology factors make ATM less interesting. The problem here is that long-distance calling isn’t nearly as profitable as it used to be, and while there are more than 12,000 local-exchange Class 5 switches, there’s only about one-tenth as many toll Class 4s that could be replaced. There seems to be almost that many voice-over-IP vendors! If all of them bet on this space, there’s going to be a high body count.

We’re not going to have a hundred-billion-dollar voice-over-IP market, but it’s not going to be totally zero, either. In sectors such as gateway controllers, prospects might actually get brighter if new custom calling features begin to drive voice competition, because these features might be easier to create on the newer gateway platforms than on traditional voice switches. Smart voice-over-IP players are looking hard at one or more of these trends, and those smart players may be able to develop a survivable niche for themselves over the next two years. Not-so-smart ones are chaining themselves to the convergence anchor.

Nolle is president of CIMI Corp., a technology assessment firm in Voorhees, N.J. He can be reached at (856) 753-0004 or