Google Advocates Unlicensed Spectrum Sharing via TV White Spaces for Wireless Broadband Access

Alan Norman, Principal of Google’s Access Strategy group, presented the company’s plans for wireless broadband using white spaces at a Nov 2nd Wireless Symposium sponsored by Joint Venture Silicon Valley. Mr. Norman joined Google’s Access Strategy team in 2010 and has since worked on a variety of projects and initiatives.

Google’s goal is to support robust, affordable and open Internet access – via both wireless and wireline technologies.  The company has been running the free WiFi network in Mt View, CA for several years,  is building out fiber to residential customers in Kansas City for 1 G bit/sec access, and (as described below) is involved in a White Spaces wireless broadband trial in Capetown, South Africa.

There are 4 ways to improve wireless network capacity, according to Mr. Norman:

  1. Use a better wireless broadband technology, e.g. 3G–>4G, higher speed WiFi (i.e. IEEE 802.11n or 802.11ac).
  2. Acquire more spectrum (but that can be very costly).
  3. Use smaller cells (macro–>nano or pico cells) with backhaul transport for each.  Smaller cells facilitate spectrum re-use within a given geographical area.
  4. Offload cellular traffic to WiFi hot spots (which use different spectrum than 3G/4G cellular networks).

Google’s view of how to achieve robust, affordable and open wireless access:

  • Support use of a balanced allocation of licensed and unlicensed spectrum
  • Advocate unlicensed spectrum sharing using  a dynamic spectrum management scheme
  • Wherever possible, support a shared wireless infrastructure (especially real estate, conduit and fiber back-haul)

Unlicensed spectrum is important because it enables:

  • Greater economies of scale in the number of wireless network endpoints and types of devices
  • Consumer identification, which results in increased investment in infrastructure
  • Lower barriers for new network operators (and new services)
  • Active competition and innovation across the value chain

Mr. Norman talked about use of TV white spaces as an example of dynamic spectrum access and spectrum sharing.  He said the U.S. has a lot of TV White Spaces (6 MHz unused TV Channels) that could be used for unlicensed wireless broadband access.  Google wants to demonstrate that over the- air TV and wireless broadband using White Spaces can co-exist with licensed spectrum.  This, in turn, would create economic opportunities and more available wireless broadband access in the U.S.

The Capetown TV White Spaces Trial:

A White Spaces wireless broadband trial in Capetown, South Africa was cited as an example of unlicensed spectrum sharing.  Alan said that 10 white space channels will be available for wireless broadband access with each one delivering 2M bit/sec Internet access to schools over 10Km distances from the carrier PoP (Point of Presence).

We later did some research on this trial and found the following information:

A group of partners, including Google, TENET, CSIR Meraka Institute, eSchools Network and WAPA, has formed to run South Africa’s first TV White Spaces trial (TVWS trial). This wireless broadband network will make use of spectrum which has been allocated to broadcasting services in South Africa, but which is not currently used (hence the term TV white spaces).

The objective of the TVWS Trial network is to provide a fast and reliable connection to the schools identified.  The trial is being overseen by ICASA- the communications regulator of South Africa.

With ICASA’s support, this group will plan, execute and report on the results of a TVWS trial to 10 schools in Cape Town in order to:

  1. Demonstrate that TVWS can be used to deliver affordable broadband and provide important Internet services without interfering with TV reception
  2. Dramatically increase awareness of the potential for TVWS technology in South Africa and on the continent more generally.
  3. The Trial Network – The trial will be conducted across a TVWS Trial Network which will include a Base Station (high site or BS) and approximately ten schools (Trial Sites) located within a 10km perimeter around the high site.

This TVWS trial network is expected to launch in December, 2012, according to Arno Hart, TENET‘s project manager for the trial.

For more information, see:

Google’s goal in this and other experimental projects is to demonstrate that TV broadcasts and wireless broadband can co-exist with licensed spectrum. They also want to show that wireless broadband using white space frequencies will not interfere with TV reception.

Proposal to Share Mounting Spaces, Power and Fiber Back-Haul Facilities:

Mr. Norman said that the cost of telecom equipment is tracking Moore’s law for semiconductors.  However, the cost of construction, civil works, conduit for fiber cables, etc continues to increase.   This has an adverse impact on fiber optic based wireless back-haul

One solution would be to share the conduit or fiber (presumably different carriers would get different fiber cables within the same conduit or use different wavelengths over the same fiber cable via DWDM).  Alan said that power for base stations/wireless access points mounted on poles or street lights could also be shared.

The bottom line here is that sharing back-haul facilities among multiple wireless broadband access providers, including light pole sharing for mounting different wireless network nodes there, would lower barriers to entry and free up capital for network growth.

Call for White Space Wireless Broadband in the U.S.:

Figure 1, heat map showing number of channels available for white space communications. Rural areas have the greatest availability. Image courtesy of Google.

Mr. Norman summed up by calling for White Space based wireless broadband in the U.S.  He said that there are plenty of unused TV channels in the 400M to 698MHz band.  He hinted that with less than five percent (5%) of the population watching over-the-air TV over only a few channels, that as much as 250 MHz of spectrum might be freed up for wireless broadband.

After his talk, I asked Alan what technology Google proposes to use for White Space based wireless broadband, considering that the standard IEEE commissioned for that is a commercial failure (802.22) with little or no global deployments.  He cited IEEE 802.11af for WiFi sharing and said there was work going on in 3GPP for spectrum sharing using LTE (that’s news to me!).

AW Comment:  Wireless broadband via TV white spaces would be particularly useful for deployment of rural broadband, because it would give rural residents broadband wireless access in areas that U.S. carriers feel aren’t densely populated enough to justify a 3G or LTE build-out.

According to the UK Telegram,

“Google and Microsoft are expected to launch a major charm offensive to win control of the valuable airwaves. They could also use the white spaces to provide widespread broadband access, potentially giving them an important calling card with which to win over customers. Google could use the white spaces as a way of offering free wi-fi services to customers with phones powered by its Android operating system, as well as to other members of the public.”

For more information, see:

Investment focused Forbes magazine recently weighed in on the rumored Google-Microsoft White Space broadband push in the U.K.

“Could it be in fact that Google and Microsoft also see a way to augment their ability to control infrastructure? And will CISCO be forced to make move into a more consumer-oriented mode? In effect, White Spaces look like a new broadband mobile system.”

For more information, see:

Muni WiFi access in Kansas City’s Google Fiber Neighborhood:

During the Q and A, Mr Norman said that there was a muni-WiFi component to Google’s Kansas City FTTH trial.  We did some research on this capability and came up with the following:

The Kansas City Star reports that, in promotional fliers for the Google Fiber service, the company said it would also install “Google-powered WiFi hotspots in your favorite public spaces around Kansas City.”  A spokesperson for the company said that the WiFi hotspots were part of the agreement with both cities for the initial Google Fiber launch.

Google will install, at their expense, WiFi antennas at over 400 locations in Kansas City. While the WiFi can be set up to be free for the public to use, Google’s spokesperson said, “It’s really up to the discretion of the public building managers.”

For more information see:


A more complete summary of the Nov 2nd Wireless Symposium, including keynote presentations from Leon Beauchman, Director of the Wireless Communications Initiative at Joint Venture Silicon Valley, and Congresswoman Anna Eschoo is at:

[Editor’s note: The heat map, in figure 1, showing channel availability was added to this article on 11/15. Thank you Mr. Norman for sharing this important visual aid.]

0 thoughts on “Google Advocates Unlicensed Spectrum Sharing via TV White Spaces for Wireless Broadband Access

  1. Thanks Alan for reporting on Google’s push for white spaces. There is so much to comment on in this post.

    I am curious if Mr. Norman talked about Google’s role as a an official FCC-approved white spaces database administrator. Their model of providing that service without user fees seems consistent with the tenor of Mr. Norman’s presentation.

    The South Africa trial is interesting, as it seems like Africa in general would be much more of a greenfield opportunity as compared to more developed markets. It seems like this would be a great test case of the idea of using technology to dynamically allocate spectrum, instead of auctions or licenses.

    As you point out, in the U.S., it seems like there could be great opportunities for white spaces in rural areas. In fact, Darrin Mylet of Adaptrum indicated as much when he demonstrated a working system in Las Vegas last year:

    They have since gone to commercial trial in rural Virginia:

    When I looked at the Spectrum Bridge database (one of the FCC white space database administrators) and entered a couple of zip codes for an article I wrote last year, the amount of spectrum varied greatly depending upon location; for instance, a sparsely populated area of Wyoming can have 35 channels or 210 MHz of availability while an urban area like Kansas City, Mo. might have just three or four channels.

    When I penned the article on white spaces in June 2011, there was some talk of production white space equipment in 2012 and perhaps in 2013 or 2014, the technology would make it to consumer devices.

    Did Mr. Norman give any hints as to when we would see white spaces move beyond commercial trials to production?

    1. Ken, Thanks for your comments. No, Mr Norman didn’t say anything more than I wrote, e.g. white spaces moving to commercial offerings. There are 2 key issues:
      1. Which standard to use for dynamic spectrum assignment, esp for LTE? Is IEEE 802.22 still a possibility? Or has the market shifted to IEEE 802.11af for WiFi spectrum sharing?
      2. Will regulators grant wireless network operators the right to use the White Spaces? Only progress on this front is the proposed FCC Incentive Auction. The vacated TV spectrum is part of White Spaces, according to Mr. Norman.

      I wrote about this recently:

      1. I believe that 802.11y is intended to address the detection of licensed use for spectrum assignment. While 802.11y specifically addresses 3650-3700 band, the mechanisms could easily be adapted to different bands. The IEEE task group chair told me this was the intent back in 2006.

    2. For the full story of the use of white space (US Navy Purchase orders, FCC before year 2000) and sharing concept with space and ground stations please see this article on Super WiFi in U.S. (use Google Toolbiar to translate to English):

      Among other claims, the ZDNET article states, “What is amazing is that the basis of its technology was invented 25 years ago, and in France!” It’s all about Hypercable and its founder Jean-Claude Ducasse

      1. Thank you Tostemcurios for the the comment. After reviewing the attached link, as well as other associated links, it looks like Hypercable has had success in bringing multichannel video via microwave frequencies to rural parts of the world and has also had success in creating heterogeneous networks many years ago.

        It sounds like Hypercable faced similar questions of interference when using the same frequencies used by Direct Broadcast Satellite.

        On a somewhat related note, it will be interesting to see what happens to US LMDS frequencies. The FCC has recently been denying extensions of U.S. LMDS licenses (they were supposed to be built out years ago).

        1. Sorry, I see now, LMDS is an old story and this 26 Ghz to 30 Ghz technology is the killer of most companies, LMDS operators in Europe and LMDS makers such as P-Com in the Silicon Valley ( P-Com are shareholder in my company in the past).Now we use this frequencies to build Multi Gigabit meshed networks for National Police..

        2. Hi, Happy new year 2013.
          We have great distance behind many concrete walls and underground penetration with our Rake system in UHF frequencies our rake is a Mix of 802.20 iburst and 802.22 ( super Wifi) based OFDM TDMA wired and ACM, test in july in Paris on You tube /
          And fortunately Purchase orders by Police for live video IP full Duplex in Mobile High speed vehicles.

  2. Great article. Indeed, White Space policy, technology and real world implementations are gaining traction globally. Adaptrum developed real technology for this space, positive 3rd party test and measurements in the UK with BBC and Arquiva, FCC certification of its CRS-1 product and many other real world deployments. We are close to launching our CRS-2 product which is made for volume commercial deployments where there are rules or willing regulators to allow experimentation etc.

    There needs to be a way for both incumbents and new entrants alike to use the fallow spectrum without causing harmful interference to others. This is great policy as it allows spectrum to be put to use. Most recent spectrum measurements indicate most of the spectrum is not used most of the time which is a real issue for consumers who want real affordable broadband.

    To answer your questions: 1. Adaptrum has a standards strategy, we are working with the White Space Alliance. 802.11af appears to be for more of an indoor type application…
    2. The more innovative regulators we have been working with around the world see white space as “Freedom Spectrum.” i.e. They are not picking winners or losers but creating a framework for broadband infrastructure to be constructed, period. The FCC IA is a topic of discussion in and of itself. In brief, the US leads the world in giving out commercial spectrum but we use the least….Taking away more open spectrum like white space when Wi-Fi is saving the carriers BILLIONS is difficult for me to really understand. It is really disappointing and not in the best interest of consumers. That being said, I think it fails anyway.

    1. To echo a key concern for white space wireless broadband commercial deployment: Which standard? If IEEE 802.11af is “for more of an indoor type application” and IEEE 802.22 WRAN is a commercial failure, then what will be used to ensure interoperability and facilitate economies of scale through volume production of equipment, components/moduels, chips, etc????
      Note also that no “alliance” is a standards making body, e.g. White Space Alliance. Why isn’t ITU-R involved as part of their 4G work?

  3. Related Google Project: Homes in Kansas City, Kansas are getting connected to the ultrafast Google Fiber service- nearly three years after Google announced the pilot project for much higher speed Internet access. The city’s Hanover Heights neighborhood is the first “fiberhood” to have the new service.

  4. Alan’s article is very informative. It was interesting to note there is a formal movement to offer services via the White Space in the TV spectrum. Admittedly I am not totally familiar with specifics of unused and grossly underused parts of the whole TV spectrum the terrestrial broadcasters, by and large, have exercised poor judgments for both programming and public interest.

  5. Another key issue for commercial deployment of White Space based wireless broadband is to conclusively prove that it won’t cause interference to digital television (DTV) signals. Even if only a small % of the U.S. public watches over the air free DTV, it’s signal quality should not be compromised by interference from unlicensed wireless transmitters operating in the White Space bands.
    Thanks for this excellent article and follow on comments.

    1. Prabhu, Thanks for your comment. Does the US White Spaces Channel Availability figure imply that interference with DTV signals will be a problem or not? I haven’t a clue! Perhaps Ken or Alan Norman can explain it.

Leave a Reply

Your email address will not be published. Required fields are marked *

I accept that my given data and my IP address is sent to a server in the USA only for the purpose of spam prevention through the Akismet program.More information on Akismet and GDPR.

This site uses Akismet to reduce spam. Learn how your comment data is processed.