Hands Off My Photons

My first paper discussed ideas about the future of a truly secure telecommunications network for mass consumption. However, it focused on traditional person-to-person voice communication (i.e. voice telephony). New technology and new forms of communication and entertainment have created demand for greater bandwidth. One possible source for this bandwidth is the electromagnetic spectrum, a shared and heavily regulated resource.

*Is current technology sufficient to create enough bandwidth for everyone to share? How much can be done within the existing regulations?

Even within the existing, tightly controlled regime, there is room for unlicensed operation of RF-emitting devices. 47 C.F.R. § 15 allows for unlicensed, low power operation of such devices within certain frequency ranges, including ranges commonly used for home wireless routers, wireless telephones, baby monitors, and garage door openers. Power for 802.11a and 802.11g transmitters is limited to 1500 W, but routers sold for home use do not even begin to approach this limit. Linksys’ popular WRT54G? router, for example, transmits at a power from 60 to 80 mW. Nonetheless, the city of Vienna is able to maintain a free wireless mesh network using about 400 nodes (many using free software modifications to the aforementioned router) that cover approximately 1/3 of the city.

However, municipal WiFi? has not really caught on in the United States. The current spectrum and power limitations mean that transmitters need to be fairly closely packed (e.g. 127 routers in the City of London to cover one square mile). Because of the relatively high frequency at which low-power consumer equipment operates, WiFi? signals do a poor job penetrating walls and roofs.

*Does cognitive radio really obviate any justification for spectrum regulation?

A better technical solution would be one that employs a longer wavelength that does a better job penetrating walls and bending around buildings. Fortunately, just such an area of the spectrum happens to be unoccupied. Termed “white spaces”, the areas of the spectrum in between television channels could be used for wireless data transmission that is ten times faster than traditional wireless routers with far greater effective range. The FCC voted to allow unlicensed use of this area of the spectrum in November, 2008. Their reasoning was that new devices could sense whether or not a given wavelength was occupied and adjust their transmission to avoid interference, termed cognitive radio. One implementation of this idea is the IEEE 802.22 standard , intended for Wireless Regional Area Networks (WRANs).

Official opening of the spectrum was expected to coincide with the transition to digital television, on February 18th, 2009. That transition was delayed by a few months, but ultimately took place successfully. However, some exceptions remain for low power television broadcasts. A search for white spaces in Morningside Heights revealed only one channel’s worth of white space: http://www.showmywhitespace.com/. The National Association of Broadcasters filed suit to overturn the FCC’s ruling under the Administrative Procedure Act just before the rule was scheduled to take effect. Given the deferential standard of review, they are unlikely to prevail.

Currently, the real delays seem to be technical instead of legal. The FCC requires the maintenance of a geolocation database within the devices to prevent their operation near recording studios and other registered venues that employ potentially conflicting devices. Further, the devices all should ideally collaborate to establish a wireless mesh capable of channel hopping if it detects interference, which is a tall order, technically, though far from intractable. As the article notes, it may be impossible to implement white-fi without the geolocation under the existing regime, since even a single packet can interfere with a concert, performance, or sporting event. Google has offered to help solve the geolocation problem, but it seems clear that the longer-term optimal solution would be to reallocate spectrum in favor of the general public and away from event-quality microphone manufacturers and other vested interests. While this seems like it may not require legislation, the FCC is equally subject to capture as a legislature, even if it's less dysfunctional.

*Is there any support in precedent for the idea that broadcast=speech, and such restrictions should be subject to a (much higher) 1st A test?

When courts scrutinize regulation of broadcast media, the degree of scrutiny depends on the nature of the regulation. If the regulation is not content-neutral, then strict scrutiny applies, and the regulation must be narrowly tailored to advance a compelling state interest. If the regulation is content-neutral, however, the regulation is subject to intermediate scrutiny under O'Brien and passes if it "furthers an important or substantial governmental interest; if the governmental interest is unrelated to the suppression of free expression; and if the incidental restriction on alleged First Amendment freedoms is no greater than is essential to the furtherance of that interest." United States v. O'Brien, 391 U.S. 367, 377 (1968).

If my local CBS affiliate's broadcast of Everybody Loves Raymond is speech, then surely my use of the internet to download that same episode must also be speech, and worthy of the same protections. If cognitive radio technology advances to the point where the devices can guarantee non-interference through a technological means, then the entire scheme of spectrum allocation is suspect. The very premise of federal spectrum regulation is set forth in Red Lion Broadcasting Co. v. FCC as there being more people who want to use the airwaves than there are airwaves to go around. 395 U.S. 367 (1969). But soon we will live in a time of plenty. The only regulation that would seem to pass the O'Brien test in that state of the world would be one that made sure that interference-generating devices weren't operated. Spectrum allocation would lack the fig leaf to cover the bare economic dictat that General Electric paid a fortune in a spectrum auction (and to legislators and lobbyists) to speak, and you're not allowed to talk back.

It seems to me that this is an unnecessarily confusing way to put the argument. Licensing speech is presumptively unconstitutional activity for the federal government, and only the technological necessity sustains it.

One of the pleasant phenomena designed into the Internet is heterogeneity. You don't need to be right about the virtues of white-fi, or any other wireless technology. We will always have the spectrum, and we'll always be able to use anything that works anywhere it works because in the end all we ever need is a way to take the next hop. You might want to get above the level of the radio protocols, or even the interests of particular incumbents, by looking again at the map of our spectrum. There's scarcely a relevant neighborhood in which the government has not reserved itself a couple of choice lots. There's always something it can either give or trade with. The real issue isn't the incumbents' spectrum, it's their incumbency. As businesses, they resent the government's competing with them in the public interest, and that resentment can be effectively brought to bear.

So leave them out of this, for a moment, and ask whether your interest is in the design of that network from a policy point of view, or in the politics of making it possible to build such a network. Then the essay can gain some focus and hit the target more neatly.