by Mark W. Hibben
Your Data Rate May Vary
What has been variously called “super WiFi” and “WiFi on steroids” by FCC Chairman Julius Genachowski may receive final approval at on upcoming meeting of the FCC on September 23. This plan, captured in a series of rules changes, calls for unlicensed “WiFi-like” devices to use portions of the TV broadcast spectrum that are unused in a given region. The amount of spectrum, and therefore data bandwidth available, varies by region, depending on how many TV stations are broadcasting to a given locale. Consumers in large metropolitan areas may have little reason to switch from their current 802.11n WiFi systems, since the new TV band devices will have less data rate capacity (lower megabits/sec). The main beneficiaries of TV band WiFi appear to be those in remote, rural areas with little or no terrestrial broadcast service, since a principal application for the new FCC rules is intended to be wireless internet service. TV band Regional Area Networks (RAN) could bring broadband internet to many who otherwise might never receive wired or fiber optic service.
Garbled in Transmission
The Tech Media could hardly contain their glee at the prospect of a whole new class of consumer gadgets to fawn over and happily regurgitated Genachowski’s hyperbole. John Sutter in CNN.com quoted him as saying, "Super Wi-Fi has extraordinary potential. What's exciting is we have a new platform for innovation and we can't anticipate what will happen next."
Sutter further states that “The changes would come about if the FCC votes to open up some of the "white space" between TV channels for use by anyone.” White spaces between TV channels? Hmm. The Tech Media seemed to regard the concept as utterly self-evident as in this description by Chloe Albanesius in PCMag.com: “What are white spaces? When TV stations switched from analog to digital signals in June 2009, it produced unused spectrum – or "white spaces" – between the TV channels. . .” Wailin Wong, writing in ChicagoTribune.com, stated “White spaces are vacant frequencies that reside between TV channels.” I’m still not sure what these writers think the term white space means, but the statements certainly imply that they think there is unused spectrum available between TV channels as defined by the FCC. The FCC and their industry partners have a very different and crisper definition of white space: white space is just the sum of the unused TV channels for a given locale. In rural areas with no broadcast TV service, this amounts to a large block of unused spectrum. It’s the recognition of this fact of spectrum underutilization in the TV bands that motivated the FCC to propose rules changes to permit unlicensed TV band devices back in May 2004. As we all know, digital communication spectrum is at something of a premium.
Overall, white space, as defined as unused TV channels, has been reduced for most areas with the advent of digital television in the US. At the same time that digital television was finalized in the US in June 2009, the FCC abolished channels 52-69 and reallocated the spectrum for other uses. This was white space for most areas of the US, since these channels were rarely used. Furthermore, digital TV in the US didn’t create or enlarge any so-called white space between channels, since there isn’t any. One of the major, unheralded accomplishments of the Advanced Television Standards Committee (ATSC) that developed the digital TV standards (usually just called ATSC standards) in the US was that they managed to fit the higher data bandwidth required for high definition video transmission into the narrow little slivers of analog radio spectrum that had been allocated to each TV channel under the old system. In the old NTSC system, each TV channel was allocated 6 MHz of bandwidth, and for the most part, these channels were contiguous. For instance, channel 8 covers 180-186 MHz and channel 9 covers 186-192 MHz, and so on. In the new ATSC system, each channel is allocated exactly 6 MHz and occupies the same frequency limits as before. This was all done at the behest of the broadcasters who wanted to preserve as much backward compatibility with the old system as possible.
The difference is that now there are no TV channels above channel 51, and the ATSC standard calls for the use of advanced modulation techniques within the 6 MHz channel band to achieve a remarkable 19.39 Mbits/sec digital data rate per channel.
In large urban areas with multiple TV broadcasters, white space can take the form of unused TV channels that are interleaved with used channels. So in a particular area, one might have only the odd numbered channels 5, 7, 9 and 11 used, with 6, 8, and 10 unused. This was mandated by the FCC to preclude interference between high power TV stations serving the same region, and the requirement continues under the ATSC standards. This may be what was meant by the “white space between channels” statements.
ATSC Leads the Way
The ATSC standards and resulting hardware implementations showed that substantial digital data transmission rates could be achieved over relatively low frequency radio waves compared with 802.11n WiFi that operates at 2.4 and 5 GHz. Thus a digital communications device such as a WiFi-like base station could achieve respectable if not great data bandwidth by using multiple TV channels if they were unused. Furthermore, because of the relatively low radio frequencies of the TV bands, such systems would be able to operate at greater ranges for the same power because of low loss due to the atmosphere and common building materials such as wood, plaster or brick.
What kind of aggregate data bandwidth could a TV band WiFi system achieve? If we take the ATSC value of roughly 20 Mb/sec/TV channel as a useful benchmark, then the total data bandwidth available to be parsed out would be just 20 Mb/sec * number of unused TV channels, subject to other restrictions that the FCC is likely to levy. The FCC intends to restrict the personal (home WiFi type) devices to only use the UHF channels 21-51 with channel 37 also off-limits because it’s reserved for other uses. That gives a maximum of 30 channels and an aggregate bandwidth of about 600 Mbits/sec. This just happens to match the maximum data rate capability of 802.11n. So why would you want to “upgrade” to TV band WiFi for the same aggregate data capacity? In many cases you wouldn’t. The strength of these systems will be that they can cover a much larger area or building with fewer drop-outs. If you happen to own a large multi-acre, multi-building property in a rural area, such a system might be the ideal way to provide broadband access anywhere on your property.
For the typical urban or suburban dweller, this range advantage is meaningless, but there may still be a place in people’s homes for TV band devices for the following reasons:
1) For short range applications, the devices can operate at very low power, and thus may be ideal to embed in home-automation systems and smart devices such as wall switches or alarm systems.
2) TV band devices can co-exist with the conventional 802.11n WiFi system on a completely non-interference basis, since they operate at very different frequencies.
3) There may be applications for TV band devices that no one’s even thought of yet.
Broadcasters Drag their Feet
The National Association of Broadcasters and other groups representing the TV industry in the US have never been fans of TV band WiFi. From the beginning, they’ve argued against opening up their reserved spectrum to unlicensed devices, claiming that they could interfere with TV channels that were in use. Much of the technology development that has been sponsored by the FCC since the 2004 announcement has been in techniques for the devices to determine reliably which TV channels are unused. Three basic methods have been identified: using an on-device GPS, combined with a data base to determine unused channel availability, directly sensing used channels in the device receiver, having the information transmitted to the device from another source. In experiments performed in 2008, the direct sensing approach didn’t work very reliably in that used channels were sometimes missed and unused channels were sometimes misidentified as used. The FCC still wants to use the direct sensing approach but is relying mainly on the GPS/database approach. Commercial fixed and home TV band WiFi base stations will use the GPS/database approach. Portable TV band devices without GPS will only be usable in a client mode with the commercial or home base stations and only be able to transmit on a given TV channel when the base station says they can.
These techniques have largely quelled the objections of the TV industry, and in the latest public filing with the FCC they no longer express opposition to the TV band devices per se.
A Spectrum Rush in the Making
Bringing broadband internet access to rural areas has been a slow painful process in the US due to infrastructure costs and the lack of a large customer population to support making the investment by rural Internet Service Providers (ISPs). From the beginning, the FCC has envisaged using TV band devices to provide large area wireless internet service. For that reason, the FCC has developed a different set of standards for commercial fixed TV band transmitters so that they can operate at higher power over more available unused TV channels than the personal/mobile devices. The IEEE has been developing a standard for this type of service, 802.22, what it calls Regional Area Networks (RANs). The IEEE has also begun working on a standard for the personal TV band device, called 802.11af, but 802.22 appears to be much further along.
In effect, TV band RANs represent a very desirable evolution of the broadcast TV paradigm for the digital age. Instead of large, high power, centralized, and one-way analog TV stations, America will see the growth of thousands of small, low power, decentralized and multi-way digital TV “stations” that carry all manner of digital content including video, web pages and personal communication.
Unlike the personal TV band devices, the commercial RAN transmitters will have to register with the TV channel data base so that TV band devices that access the database know they are active. Since the commercial RAN transmitters are nevertheless unlicensed, it’s not at all clear how competition between RAN operators will be regulated. If different Wireless ISPs (WISPs) want to erect RANs in the same region, how is the unused TV spectrum apportioned between them? Right now, it appears to be a matter of “first come, first served” with whoever erects a RAN transceiver and registers first getting to use the available TV bands in that locale. This could ignite a “spectrum rush” with RAN operators hurrying to set up RANs (mostly in urban areas) just so that they can get registered in the data base first, even though a customer market for the RANs hasn’t been clearly established. I wouldn’t be surprised if the stock markets see a short-lived WISP bubble as a result.