Princeton professor of Computer Science and Public Affairs, Ed Felten, has published a ten-page whitepaper on network neutrality that greatly clarifies the complexities involved in the issue. Felten’s paper does a good job of finding the third-way between the misleading scare tactics of some activist organizations and the telecommunications companies’ astroturf campaigns.
Felten posits that the mechanics of network discrimination are the core of the issue, something that relatively few people understand.
Because the internet is an end-to-end network, with all of the intelligence and heavy lifting happening at the edges, Felten writes that the network neutrality debate is a “fight between the edges and the middle over control of the network.”
The internet has enjoyed its unparalleled success precisely because all intelligence has been pushed to the edges of the network. The telecommunications companies want to reverse this, aggregating the intelligence in the center of the network, where everything—including innovation—is easier to control. Putting the internet genie back in the bottle, as it were.
Network discrimination can occur, according to Felten, when a router prioritizes packets and discards low-priority packets first. Felten calls this minimal discrimination because packets are only discarded when it’s absolutely necessary.
But if packets are prioritized, routers can be configured to discard low-priority packets even when it’s capable of delivering them. Felten calls this non-minimal discrimination.
It’s important to allow minimal discrimination—not all packets can be passed all the time, but every router makes a best effort—but disallow non-minimal discrimination. As Felten points out, “minimal discrimination may sometimes be an engineering necessity due to the finite speed of network links, but non-minimal discrimination is never technologically necessary.”
This best effort principle is what makes the internet possible and contributes greatly to its success.
Felten writes that network discrimination can also occur when routers delay delivery of packets by reordering them. TCP/IP doesn’t place restrictions on how or when packets are forwarded; routers can choose any packet it wants to forward next. The same minimal and non-minimal discrimination definitions apply to packet delay. Again, minimal discrimination should be allowed; non-minimal discrimination must be prohibited.
Delay discrimination can impede some applications more than others because of “jitter”—unpredictable changes in delay. Voice Over IP (VOIP) applications, for example, rely on a steady connection; jitter causes VOIP to become unusable. And the telecommunications companies would like nothing better than to make other vendors’ internet telephony applications unusable. “Causing jitter for such services,” writes Felten, “whether by minimal or non-minimal delay discrimination, could be an effective tactic for an ISP that wants to drive customers away from independent Internet telephone services.”
Felten provides no policy recommendations other than to emphasize that we should be concerned about unintended consequences. The current policy—network neutrality is on everyone’s radar but no regulations are yet entrenched—may be the best situation for now, according to Felten. The telcos are unlikely to draw the attention of regulators by rampant packet discrimination, and the headaches associated with fixing and enforcing possibly bad legislation are avoided. Sometimes political gridlock is good.
0 responses. Comments closed for this article.