Oddly Enough, Access Is Where QoS Really Helps

Brough Turner, NMS Communications CTO, argues that quality of service measures in the Internet backbone provide negligible benefits. It's in the access links where QoS really can make a difference, and that's the area where at&t's merger approval agreement with BellSouth prohibits it from doing so. Oddly enough, the inability to "discriminate" between packets prevents users from experience improvements they might like to have.

Once packets get beyond the access network, every link in the Internet is carrying multiplexed traffic that is statistically relatively predictable, Turner argues. "Traffic volumes vary by time of day, but these links don't saturate, except as a result of poor engineering or forecasting on the part of the ISP or failures in other parts of the network causing rerouted traffic," says Turner. "Either case generates a rapid response from any ISP that expects to remain in business."

"In short, except at the edges of the access network, Internet links may be heavily loaded but are not saturated," says Turner. "Best effort is good enough even for low latency applications like voice telephony" in the backbone, as a result. "Except during major failures, the effect of QoS in the Internet backbone is negligible."

That isn't the case for access links, Turner argues. "There is one place in the public Internet where limited, highly specific QoS measures make sense and are being deployed," and that is "at the consumer end of an asymmetric broadband access link."
Specifically, it is the upstream links that can saturate. The other issue is that it frequently is not possible to buy additional capacity in the upstream direction, at any price. And that makes routing policies very valuable.

His full comments can be viewed at http://blogs.nmss.com/communications/.