Why free spectrum doesn't mean free Wi-Fi

A recent Washington Post article about the Federal Communications Commission's goal of making parts of the 600MHz spectrum band available for use free of charge was written in a way that led some readers to believe that the FCC was proposing a nationwide free “super” Wi-Fi network.

Diligent tech writers and bloggers quickly pointed out how the article was actually confusing several topics – FCC auction of 600MHz TV spectrum, allocation of more 5GHz spectrum for Wi-Fi, and TV white spaces – and coming up with a grand vision of nationwide free Internet access. Rather than adding to the chorus of criticism of the article, let's revisit the notion of a nationwide free Internet network and the fallacy that free spectrum equates to free service, along with some of the other well-worn ideas that tend to travel with this topic.

Yes the spectrum is free, but not the rest of the network

No question about it, spectrum is a key resource for building a network. It can also be a very expensive asset, especially in the US where the Federal government sees spectrum auctions as an important tool to raise revenues. So the concept of free spectrum certainly gets people's attention. However, one should not confuse free spectrum with free service. Take Wi-Fi networks as an example.

Wi-Fi already uses unlicensed spectrum, and many of the networks that use Wi-Fi gear charge in one way or another for the service, such as with AT&T's Wi-Fi network. Larger nationwide Wi-Fi networks certainly charge for Internet access either directly or indirectly, as a cost component of a larger service plan. The reason service providers must charge for Wi-Fi is that behind that spectrum lie base stations, backhaul, a core network, and the internet itself. All of those parts cost money – they don't just come with spectrum like a prize in a Cracker Jacks box. Somebody has to pay for them.

Now imagine the cost of doing that across the US. If what passes for nationwide Wi-Fi network services today can't support the service without charging, what are the chances that full nationwide coverage can be achieved without charging the end user? Given the approximately $US7 billion it has already allocated to the yet-to-be fulfilled national broadband initiative, the US government is unlikely to pick up the bill.

Please don't bother me with advertising-based service models

Most proposals of “free” communications services are backed by a business plan in which advertisers will pick up the costs of the service. I have heard this business model in the past with long-distance dial-around services, dial-up Internet access, and broadband services. But none of these services has yet to give their pay-service counterparts much of a scare. Why is that?

Well for one thing, advertising funds are not a bottomless pool of money. Companies don't have unlimited advertising budgets. Plus, what exactly is the value of spending money advertising to people who have already self-identified themselves as being miserly by using the service? These types of service models have come and gone, much like municipal Wi-Fi, while their fee-based counterparts remain. I don't see that changing anytime soon.

A “super” Wi-Fi isn't going to replace the mobile network

Another topic that comes up when talking about a nationwide free Wi-Fi network is that it will eventually allow people to cancel their cellular services. This notion goes all the way back to the initial introduction of public Wi-Fi networks. In the particular case cited in the Washington Post article, the 600MHz spectrum under consideration would allow for higher output power than is available with current Wi-Fi access points, which would give it greater range than current Wi-Fi networks. But even with greater radio signal range, this free network would still be far from commercial-grade.

Beyond the cost issues discussed earlier, to replace the cellular network this “super” Wi-Fi network would need to handle base station–to–base station handoffs to ensure seamless coverage, not to mention the ability to maintain a network session while the user moves at high speed. All are qualities not currently available with Wi-Fi. Plus the amount of spectrum being discussed at 600MHz would be subject to severe congestion: the FCC is only talking about possibly 6MHz to 18MHz of spectrum being made license-free.

This falls well short of the holdings of the big four US mobile operators, all of which own substantially more spectrum and still have congestion concerns. If AT&T has well over 100MHz of spectrum and still needs more, imagine how well a nationwide 18MHz network would hold up, especially if it costs nothing to use.

Daryl Schoolar is a principal analyst specialising in Network Infrastructure at OVUM.