editorialHandicapping the network sweepstakesLast month, I sat in a briefing room and enjoyed a very entertaining discussion with executives from one of the major communications vendors on the future of optical networks. The discussion centered on where routing intelligence would reside in the network. Say you had a data signal going from New York to Los Angeles. According to the executives, the piece of equipment originating the signal at the very edge of the network would put a final address on the signal, but that would be about it. Routing equipment in the network core would determine whether that signal reached Los Angeles via Chicago and Denver or Atlanta and Phoenix. No one, these executives assured me, would want the edge device to be responsible for determining each point in the route.The next day, I met with one of the chief technologists at an emerging optical carrier. Someday, the technologist said, edge devices will be able to route traffic from start to finish, allowing customers to pick the routes their traffic will take across the country. Naturally, I told the technologist about my conversation with the executives from the network equipment company, who had said no one would want to do what he proposed. Well, that seemed to be the direction technology was heading, the technologist replied; why not go ahead and do it? The point of this story is not that vendors don't know what they're talking about-it's that when we discuss "tomorrow's networks" or "future topologies," I sometimes think we're doing little different than your average bookie. We're really laying odds on what we think will happen. And that can include touts at carriers as well as vendors. After all, who's to say the executives mentioned above aren't right? So you lay your bets and you cross your fingers. Certainly, the articles in this month's issue of Fiber Exchange illustrate this concept. Passive optical networks (PONs) have suddenly become one of the hottest areas for startups and established vendors alike. But only a year ago, fiber-to-the-business and -home was considered little more than a pipe dream by most prognosticators. Who knew? More to the point, who knows now? The same can be said for the great debate over mesh versus ring topologies. Mesh certainly makes sense for data, and traffic patterns are swinging increasingly toward data, as the Internet and other new services come online. Yet voice continues to bring in the most revenue. So what do you do-build for traffic or build for revenue? We're all just gambling here. And no one wants to go double or nothing. Got your rabbit's foot handy? Stephen Hardy Editorial Director & Associate Publisher |