Monday, May 22, 2006

Net Neutrality: is Tiering reasonable?


The carriers desperately want to support tiering of network services, ostensibly to provide video services to consumers. The big financial win for the telcos is to supplant cable providers as the premier provider of television services. Using a tiered delivery system, IPTV promises to offer a rich, lustrous landscape of integrated Internet, television, and telephony.

But this is a very different approach to the public Internet as it has historically operated. All packets are treated equally in the current network architecture. Each is given the best treatment available ("best effort"), but no one packet is given special priority over any other. Thus, a video packet is treated the same as a chunk of a web page, or a piece of a streaming radio broadcast. Packet discrimination is forbidden, whether based upon the type of packet or the source of the information. That is the concept generally termed, "net neutrality": all packets are created equal.

The carriers tell us that their new, advanced services require "tiering", or segregation of packets based upon priority. In this scheme, each packet must be inspected to determine its raison d'etre  -- its reason for being. With this model, important packets are accelerated through the system, while lesser packets are permitted to queue. An HDTV packet, for instance, might get high-quality treatment ("QoS" or quality-of-service) while a routine web server response could get second-class treatment.

That sounds entirely reasonable. But there are at least a couple of important questions that we must ask ourselves:

1) Will tiering work (and has it ever worked)?
2) What are the risks of allowing the carriers to tier without regulation?

Will tiering work? Has it ever worked?


The best example of a next-generation, high-speed network is probably Internet2's Abilene network. Abilene, supported by a consortium of industry and academic institutions, provides a backbone that supports 10 gigabits per second and a stated goal of offering up to 100 megabits per second between every workstation. Stated goals of the network are to create revolutionary new Internet applications and to disseminate new services and applications to the broader Internet community.

With 207 participating Universities, a fiber backbone provided by Qwest, hardware from Cisco, and many other industry participants, the Abilene network should have been a perfect test-case for tiering. In fact, Internet2's QoS working group was formed years ago to advance network applications through IP traffic differentiation.

What happened? Internet2's QoS and tiering efforts failed:

The utter simplicity of the best-effort service model is one of the essential reasons for the success of the Internet. It has allowed IP to be implemented over every conceivable link-layer and has... enabled a fast, dumb, cheap, and wildly scalable internet.

The greatest challenge for internet QoS is to find lightweight traffic differentiation schemes that add value without adding significant additional operational complexity or endangering the principles that have made the Internet so successful...

...After several years of experience attempting to deploy an interdomain, EF-based, virtual wire service in the Internet2 environment, the Internet2 QoS working group has concluded that any reservation-based form of QoS faces prohibitively difficult deployment obstacles...

Put simply, the best and brightest -- culled from industry and academia -- failed to implement a QoS/tiering scheme.

Wouldn't it make sense to get tiering to work in a controlled environment first, before trying to deploy it on a much more diverse and challenging Internet landscape?

What are the risks of allowing the carriers to tier without regulation?


If there were competition for last-mile services to consumers, one could argue that permitting packet discrimination might make sense. But 98% of Americans have two or less choices for broadband. The potential for abuse is therefore present, according to Rep. James Sensenbrenner (R-WI), who notes, "...this virtual duopoly creates an environment that is ripe for anti-competitive abuses."

What evidence do we have for that? For one, the history of the carriers' behavior related to net neutrality is not proud. Columbia's Tim Wu points out that before net neutrality was enunciated and enforced by the FCC:

* AT&T warned customers that using wireless home networking equipment was a 'federal crime'.

* Cox Cable disciplined users of virtual private networks.

* Comcast blocked Internet VPN ports, which prevented Washington state workers from telecommuting.

* In 2004, North Carolina ISP Madison River blocked their DSL customers from using any rival Web-based phone service until disciplined by the FCC.

* In 2005, Canada's Telus blocked customers from visiting a Web site sympathetic to the Telecommunications Workers Union during a labor dispute.

* Also in Canada, Shaw Communications levied a $10-per-month tarriff on the Vonage telephone service, which competes with Shaw's own VoIP offering.

If we revisit the COPE Act, which is the legislation the carriers are backing, the words block, impair, and degrade are nowhere to be found.

Yet, AT&T's policy website states:

We will not block, impair or degrade access to any legal web site, application or service, nor will we intentionally degrade the customer experience or the service delivery of content or application providers.

We must ask: why are those very important words omitted from the COPE Act?

The FCC must be empowered to instantly detect and eradicate this type of anti-competitive behavior on the part of the carriers. Otherwise, the entire Internet revolution may be endangered.

Conclusions


Let's ignore, for the moment, the failures of tiering in controlled network situations such as Abilene.

When the carriers -- and not consumers -- get to discriminate between packets, what happens to ventured-funded innovation, startups, and America's technological leadership position?

The two guys in a garage — who can’t afford to pay the tiering tarriff to each and every telco and cable company — won’t invent the next  Google.

And those four nerds down the street, the ones with a brand-spanking new AJAX-enabled Web 2.0 auction service? Well, they’ll never get any venture funding because eBay can afford to pay the tarriffs and effectively block upstarts.

Oh, and that peer-to-peer company that came up with a new VoIP alternative to challenge Vonage and Skype? Well, they’re dead in the water because that treads on the telcos’ turf.

Until there is true competition for last-mile services, the carriers should be forced to abide by network neutrality. And let's let Abilene or a similar controlled, experimental network prove that tiering is feasible... before unleashing it on the value-creation machine that is today's Internet.

Go to Save The Internet today and take action.

No comments: