Do Neutral Wireless Networks Require an End to the Flat-rate Plan?


The network neutrality debate — whether or not Internet Service Providers can discriminate against packets (GigaOM Pro sub req’d) or application providers — pits what the blogosphere often sees as the forces of good (Google, The Free Press) against the forces of evil (AT&T, Verizon, Comcast), while generally ignoring the technical realities or even clearly understanding the limits of said networks. So I was excited to see presented to the FCC this week a paper written by Scott Jordan, a professor of computer science at the University of California, Irvine, on whether or not one can or should apply net neutrality to wireless networks.

The paper concludes that the differences between wireline and wireless networks do change the way network management is implemented, and suggests that by creating the equivalent of an open interface for the transport layers (layers 1-3 in the OSI model) of a wireless network would be enough to prevent ISPs from stifling competition on wireless networks. From an abstract of the paper:

We address whether differences between wired and wireless network technology merit different treatment with respect to net neutrality. We are concerned with whether the challenges of wireless signals and mobility merit different traffic management techniques, and how these techniques may affect net neutrality. Although wireless networks require stronger traffic management, we find these differences are only at and below the network layer, and hence wireless broadband access providers can effectively control congestion without restricting a user’s right to run the applications of their choice.

However, for Jordan the open interface would be tied to pricing, notably the amount a user is willing to pay for certain prioritization or types of traffic at the lower levels. He explains in the paper itself:

In contrast, many current plans are not application-agnostic and are hence not consistent with an open interface. Some plans for smartphones include unlimited amounts of data, but restrict use to certain devices (e.g. prohibit tethering to a laptop) and to certain applications (e.g. permit web browsing and email, but prohibit ?le sharing, streaming, and VoIP). The goals of traffic management can be more efficiently obtained through an application-agnostic interface that allows users to choose their own applications and to match these applications to QoS options based on price.

Not only does this mean the flat-rate mobile broadband plan is dead, but the onus is on the consumer to understand what she wants to do with her device and subscribe to the correct pricing plan. Already carriers are weighing how they will change mobile broadband pricing (GigaOM Pro, sub req’d) to more accurately reflect usage, so by providing a way to offer usage-based pricing in a way that could fit with network neutrality rules, this paper could help carriers implement such plans. It explains from a technical perspective why wireless networks should also abide by network neutrality regulations and how to do so in a manner that respects the constraints unique to wireless networks. So there’s something in here for both consumers and carriers to potentially dislike.

Image courtesy of Flickr user pfly



Tom seems to have the solution when he says “the answer is mettering in the same way the DSL/cable operators have done it. keep everything unlimited and sell tiers of speed.”


Tom, when you talk about tiers and different options, you are describing the market. One of the many reasons to fight Net Neutrality is to preserve the benefits of choice that only a market can provide.

If people don’t like certain pricing models, then entrepreneur will figure out better ways.

As soon as the government is involved, it will be harder for competition to develop simply because of the uncertainty of government.

I’m surprised Gigaom is so anti-competition here.


i really believe the answer is mettering in the same way the DSL/cable operators have done it. keep everything unlimited and sell tiers of speed. the basic tier may have to be really slow compared to todays standard. but i think that may be OK. i know large families who share Cricket mobile broadband connections that are only providing about 400kbps and are happy. how about selling unlimited 0.5mbps connections @ $15/month but letting consumers bundle as many ‘connection as they desire. so while $15 gets you 0.5mbs $150/month would get you 5 mbps. most consumers would go for the cheapest option keeping the network uncontested and the power users who believe they need the speed could pay for it.

and if the operator really want to sell by the KB let them do that as well which would actually makes a lot of sense for certain application that do not need lots of constant bandwidth. but please give us a slow but unlimited option.


So someone had to right a paper to explain that there will never be enough OTA bandwidth to satisfy demand.

OK, I get it.

What we’re left is trying to come up with a market structure that follows a rational supply and demand curve, a structure that would naturally throttle demand as the supply of unused bandwidth approaches zero.

Flat-rate pricing can never do this since it’s based on the assumption that bandwidth is infinite. Carriers throttle arbitrarily whether or not it’s required. Worse of all, customers are charged whether or not they use the resource. The result is supply and demand have no relationship with each other and no inherit corrective tendencies.

Like it or not, metering is a possible solution. By charging by the byte, users would finally be aware of how much of the resource they’re using. Overall the price of a byte would be set at a level necessary to protect the network. Set too high and demand and profits go down, too low and network reliability suffers.

Whatever shape broadband pricing takes in the future, it must balance supply and demand.


i believe that there is a reality in play here which is consumers are completely unwilling to subscribe to a service with such an unpredictable cost as metered broadband would be that it simply would not work.


Keep the government out. Imagine if the government regulated PC’s. We’d all be running something like Windows ME and the people trying to run better OS’s would be taxed and ridiculed by the Politically Correct people.

Comments are closed.