January 17, 2001, 10:16 AM — It may be easy to overlook Sprint these days in light of the big shake-ups at rival carriers AT&T and WorldCom. But Bob Azzi, Sprint's vice president of engineering, says the carrier's efforts to offer bundled services based on a converged network are resonating with customers. Azzi talked about Sprint's strategy with Network World News Editor Bob Brown. What do you think of the recent restructuring announcements at AT&T and WorldCom?
They are Wall Street strategies. From our perspective, customers want a complete bundle of services. If you keep divvying up your company, it gets more difficult to provide that complete bundle. Our strategy is to have wireless, Internet, voice and data available in one place.
Speaking of service bundles, how is Sprint's Integrated On-Demand Network (ION ) service coming along?
It is the centerpiece of our convergence strategy [ION is Sprint's integrated voice, data and video service for business and residential customers]. We are in the testing phases of the ATM core of that network, which will become available in the first part of next year. 2001 will be an important year for ION, because we'll be rolling it out widely geographically and expect to see a lot of customers come online.
Why an ATM core rather than an IP one?
Right now, the voice [part of ION] is provisioned over an IP infrastructure because we could get there faster. The target architecture is an ATM core for voice to ensure the quality of service [QoS]. We do not believe IP is ready in terms of [QoS] for voice that customers want. They'll tolerate less quality if it's the only way to get to Africa or the Far East and it's cheap, but if you're having a business conversation, you want the voice to sound like the voice you're used to getting today.
When you see other carriers talking about IP and QoS, when you sniff behind it, you'll usually find ATM at the core. At some point, as ATM traffic becomes a smaller percentage of overall traffic, I could see us slamming the ATM traffic onto the IP network and giving the traffic some QoS attributes that give it an express route through the IP net. There are emerging protocols we're looking at that might be able to help us do that.
We tend to get caught up in the newest technologies, such as ION. I'm wondering how an older technology, frame relay, is holding up?
Frame relay is a growing business for us, and we're still building out our frame relay network. . . . It pays some of the bills. We expect to see some slowing of the frame relay growth over the next few years, but when customers find something that works for them, they tend to stick with it for awhile. There haven't been any real frame relay service breakthroughs of late; it's a fairly mature technology.