Network Management News from All Corners

The past week has seen several important developments relating to how network operators manage the traffic on their networks. After several months of discussions, the Internet Engineering Task Force (IETF) – one of the leading technical standards organizations for the Internet – officially chartered a new working group focusing on Congestion Exposure (CONEX). The purpose of CONEX is to create a mechanism that can inform network nodes about the congestion that packets are encountering on the network. As we explained around the time of the last IETF meeting,

“The idea behind CONEX is simple: to give network nodes (and the companies that operate them) a simple way to know and account for the volume of congestion on the network at the very times when the network is congested. The technical details are a little hairy, but the basic notion is that Internet packets would carry information about the congestion they observe in their IP headers so that any router along the path could read it.”

Although the core work items that have been set out for CONEX are “non-standards track” (meaning that they will not be directly adopted and deployed as standard Internet protocols, although that may be possible further in the future), the creation of the working group and the engineering work that it will spur represent important steps towards developing congestion management techniques that are both highly responsive to congestion as it occurs and agnostic to which applications may be causing that congestion. The latter aspect is particularly important from a policy perspective, as network operators’ use of management techniques that target particular applications has raised fears among CDT and others about ISPs serving as network gatekeepers, picking and choosing which new applications will succeed or fail. CONEX should provide a valuable foundation for building responsive, non-discriminatory network management tools.

The CONEX announcement came (perhaps somewhat ironically) only a day after AT&T announced that it would introduce new data-capped wireless plans, in contrast to its current unlimited data offerings. The two new plans offer users baseline data allowances at flat rates: 200 MB for $15/month and 2 GB for $25/month, respectively. Users who exceed these allowances will be charged for additional allowances (200 MB for $15 on the smaller plan and 1 GB for $10 on the heavier plan).

CDT has long supported this sort of usage-based pricing as long as the data caps are not based on the use of particular applications, the charges are fair, and users can easily determine how much bandwidth they’ve used. It seems logical for network operators facing high traffic volumes to charge those users who place a heavier burden on the network more for what they use.

But comparing data caps to the kinds of precise, real-time measurement and management solutions that CONEX may facilitate reveals that not all of those burdens are created equal. While data caps will certainly make users more aware of their impact on the network and may well curb heavier or more price-sensitive users’ activity, they can also be viewed as a fairly blunt instrument that may do little to mitigate instantaneous or episodic congestion.

Imagine, for example, if next week a good fraction of the nation’s soccer fans decide to stream live video of a particular World Cup match unobtrusively on their mobile devices while they’re at work, or if they all check out the video highlights of the day’s games during their commute home. This short-lived data consumption may not send too many of them over their data caps (or they may be willing to pay for it on this one occasion), but it may well create instantaneous congestion problems on the network as all of them suck down those video bits. CONEX presents the possibility of developing more fine-grained accountability mechanisms that can help to mitigate this kind of congestion as it is happening, and it could also incentivize the creation of technologies and protocols that consume bandwidth in a way that is more friendly to others using the network simultaneously. While data caps may serve an immediate need, much work remains in refining and targeting congestion management tools at the specific, real-time issues that users encounter on the network.

Those issues, and many others relating to network management, may soon be discussed and debated within the Broadband Internet Technical Advisory Group (TAG), an initiative announced today by a group of Internet technology companies. The TAG seeks to serve as a venue where technical experts can develop consensus about broadband network management practices. CDT is hopeful about the TAG’s ability to address some of the Internet’s key network management issues, although doing so would require that Internet stakeholders of all kinds — including those representing users, the public interest, applications providers, and network providers – be allowed to participate. Absent the presence of a diversity of technologists, it will be difficult for the TAG to credibly reflect the technical interests of the Internet community at large.

And while the formation of the TAG, the chartering of CONEX, and the new AT&T data plans have all been announced amidst a maelstrom of activity at the FCC relating to broadband Internet openness, none of them replace the need for the FCC to pursue its stated plans to clarify its broadband Internet jurisdiction and take action to ensure the Internet's openness and neutrality. Only the combination of technical, business, and regulatory efforts can ensure that users will continue to enjoy the Internet in its most open (and least congested) form.