Monday, October 12, 2015

Can Regulators Actually Detect Network Neutrality Infractions? Study Suggests Answer is "No"

Is it possible to detect and measure network neutrality infractions?

It is not a rhetorical question, according to Ofcom, the U.K. communications regulator. After commissioning a study on whether it is possible to create meaningful network neutrality rules, Ofcom has produced a report suggesting it is not--at present--actually possible to determine whether impermissible traffic shaping actually has occurred.

The obvious implication: how can engineers trying to enforce network neutrality rules actually ascertain that the rules are respected, when measurement is not possible?

Those are questions raised by a study conducted by Predictable Network Solutions for Ofcom, the U.K. communications regulator.

The study found that none of the existing and methods  tools actually succeed at detecting most forms of traffic management.

The reason for the ineffectiveness of the these tools is the complexity of the Internet; when there’s a delay between two endpoints, the tools are unable to pinpoint the cause of the delay.

That poses a key problem for network neutrality: there is no way, at present, to determine if an infraction has happened.

Though it always is dangerous to infer too much from any single study, on a subject as complicated as traffic management, that is potentially highly significant.

The findings cast a shadow on net neutrality regulations, which are an attempt to ban behavior that generally can’t be detected, let alone measured, says Richard Bennett, a communications consultant.

And even if the behavior could be measured, it’s not always anti-consumer.

“If I was on a business Skype call from my house while the kids were watching multiple video streams, I would like it to be differentially treated because I would like my voice call not to fail,” says Neil Davies, Predictable Network Solutions principal. “Differential management itself is not necessarily against the end user’s interest. It could be very much for it.”

Differentiated traffic “obviously has a value to the end user,” and could “potentially garner a price premium,” Davies says. That, alas, is the argument many supporters of QoS, would make.

Users should be able to choose which traffic gets priority, Davies says. Some would argue ISPs likewise should be able to offer such capabilities to their customers.

Though there will be debate about the findings, the PNS study suggests network neutrality rules are unenforceable, because application-specific discrimination isn’t detectable by any known tool, whether it’s NetPolice, NANO, DiffProbe, Glasnost, ShaperProbe, or ChkDiff.

No comments:

Many Winners and Losers from Generative AI

Perhaps there is no contradiction between low historical total factor annual productivity gains and high expected generative artificial inte...