How do technologies die? There is no announcement of their demise. Like Homer Simpson, they simply fade back into the bushes. Acronyms like ISDN and WiMAX used to dominate telecoms news headlines but today you would never know they existed. Billions of dollars have been spent on technologies that never quite fulfilled their promise. There are many reasons why technologies fail: lack of adequate investment, lack of complementarity with other market products, lack of regulatory support, bad timing, the list is long. For the supporters of a particular technology, it is often difficult to know when growth has stalled, whether more patience and investment is required, or it is time to move on because you are flogging a dead horse.
This is where I find myself with TVWS or Television White Space technology. I have been an advocate of TVWS since 2008. Fourteen years later, TVWS has seen some successes but has yet to achieve widespread adoption.
For the uninitiated, TVWS is an innovation both in wireless regulation and technology. It is a wireless broadband technology for Point-to-Point and Point-to-MultiPoint links that operates in the wireless spectrum traditionally used by television broadcasters. Historically, television broadcast technology has always relied on leaving empty channels between broadcast channels. TVWS was designed to fit inside those empty channels without interfering with the television broadcast. While the way television frequencies are organised is changing with the transition to digital broadcasting, the fact remains that, in most countries, a significant part of the frequencies allocated for television broadcast remain unused, especially in rural areas.
What’s exciting about TVWS?
Two things make TVWS a unique and interesting technology. The first has to do with the radio spectrum that TVWS uses. The propagation of radio frequencies changes as you move up and down the frequency bands. Lower radio frequencies, such as those used in television broadcast, have the virtue of being able to penetrate foliage and even navigate larger obstacles like hills, buildings, etc. Compare this with WiFi technology, operating at 2.4 or 5.8GHz, that requires a clear line of sight between radios in order to operate. In certain cases, TVWS, which uses broadcast spectrum in the range of 470-698MHz, could lower network deployment costs by only requiring a single point to point connection where a WiFi or other wireless backhaul technology would either have to introduce one or more intermediate links to get over a hill, for example, or build a much taller tower in order to be able to “see” the other radio antenna. This is known as a Non Line-of-Sight technology (NLOS). While TVWS is not infallible, it represents a very useful tool for an internet service provider to have in their connectivity toolbox.
The second, and arguably more important, issue that makes TVWS unique and interesting is that it represents a new approach to the management of radio spectrum. Wireless technologies often take a long time to come to fruition because the use of specific wireless frequencies for a particular technology is governed by national regulatory bodies who in turn are guided by a UN treaty organisation called the International Telecommunications Union (ITU) which brokers consensus on how specific radio frequencies should be used. Brokering international consensus is a slow moving process in any field and this is no less true in telecommunications. It can sometimes take a decade or more for a wireless technology to be widely recognised by national regulators. The upshot of this is that technologies often atrophy many years before consensus can be reached on the re-allocation of wireless frequencies they need in order to operate.
TVWS represents a complementary spectrum management strategy that allows for the re-use of frequencies assigned to television broadcast on the condition they do not interfere with any existing or planned broadcast signals. Kind of like letting someone stay in your house while you’re not there.
I think of TVWS as a kind of “while you’re waiting” technology by which I mean to say that it doesn’t necessarily depend on slow international cycles of frequency reallocation. It allows for a radio frequency to be used as long as nobody else is using it. TVWS was first proposed back in 2003 in the United States. The unused channels in television broadcast frequencies seemed like a perfect fit for this serendipitous re-use of spectrum.
It is important also to note that while the principle of secondary re-use of spectrum was pioneered with TVWS, the principle can be applied to any spectrum. Indeed, both the US and UK regulators have been applying this principle to LTE/5G spectrum in different ways.
The Initial Spread of TVWS
In the United States, where TVWS was pioneered, advocates of the technology encountered a lot of political opposition. Broadcasters expressed concern about potential interference with television broadcast and wireless microphone manufacturers (who were already squatting in unused television spectrum) were also concerned about potential interference. The push-back from these organisations was such that, while TVWS regulation was enacted, the constraints placed upon its operation were extremely strict, limiting its effectiveness and ease-of-deployment. Not only was TVWS very limited in terms of power output and how close it could operate to television channels in use by broadcasters, but it was also determined that TVWS could not be trusted to detect and avoid television broadcast frequencies on their own. It was determined that all TVWS devices must therefore regularly authenticate with a geo-location database which would determine which frequencies, if any, were available for use. The creation and operation of these databases was outsourced to the market and it was assumed that database service providers would make sufficient revenue from operators paying a modest sum per device to access the database. That was at a time when it was assumed that IoT was a part of the TVWS landscape and there might be millions of devices connecting to these databases, creating a competitive market for database service providers.
This is the regulatory model that was ultimately exported to the rest of the world. As a pioneer in spectrum regulation and a massive market unto itself, regulatory innovations, other countries often follow where the US leads in spectrum innovation. The UK, Singapore and a few other countries adopted TVWS and implemented geo-location databases with slight variations in frequencies and power levels in each country. On the African continent, trials in TVWS began in late 2012 in Kenya and South Africa, supported by Microsoft and Google respectively. Trials spread to Malawi, Mozambique, Nigeria, Ghana and other countries in the region but actual regulation was slow in coming. In 2013, Google and Microsoft (and others) founded an industry association called the Dynamic Spectrum Alliance to promote TVWS regulation. They developed model regulations to assist regulators in adopting TVWS.
What Happened After
In the United States, everything initially seemed to go as planned. Regulations were introduced and geo-location database providers were certified. Microsoft got behind TVWS deployments in the US with their AirBand initiative. However, the restrictive TVWS rules limited adoption and that limited demand meant that costs remained relatively high as manufacturers were unable to achieve economies of scale in production.
Meanwhile, WiFi was going through a small revolution. The discovery that low-power WiFi could be focused into a high-speed, point-to-point broadband technology unlocked a whole new industry of manufacturers producing inexpensive WiFi equipment for backhaul. Names like Mikrotik, Ubiquiti, Cambium, Mimosa and others were transforming wireless backhaul from a technology costing tens of thousands of dollars to one costing hundreds. Now, in order to survive, TVWS had to be cheaper and more effective than the WiFi industry, a task that has become increasingly difficult as WiFi equipment grows in sophistication even as it drops in cost.
In African countries, TVWS regulation has been passed in South Africa (2018), Mozambique (2019) and Kenya (2021). It has gotten as far as a consultative stage in countries like Ghana, Malawi, Nigeria, and Uganda but failed to get any further. My belief is that the challenge of how to implement a geo-location database has been the chief stumbling block for these latter countries.
The “market forces” approach of the United States, which assumed that a thriving TVWS industry would pay for the operation of a commercial geo-location database service provider, has not really found itself not fit for purpose in African countries. For communication regulators in emerging markets, the problem with geo-location databases is manifold. Who will pay for the database? Who will operate the database? Who will validate and oversee the database? Can we really trust a database? The list goes on.
Companies trying to make a business out of TVWS geo-location services have struggled. Most of the first companies to offer these services have moved on. Nominet, the .UK domain name registrar decided to get into the geo-location business but failed to make a go of it. They sold the technology to a French company called Red Technologies. Finnish company Fair Spectrum have stuck it out the longest and still offer TVWS geo-location database services.
South Africa is the exception to this thanks to the Council for Scientific and Industrial Research (CSIR) who invested significantly in developing their own TVWS geo-location database. Like others, CSIR hoped to commercialise their database technology but that has yet to transpire. The good news is that the South African regulator is the beneficiary of their hard work, being able to rely on CSIR for geo-location database services. It is not clear whether money is changing hands for that service, but I doubt it. As both entities are funded by the South African government, it might be a bit like taking money out of a left pocket to put in a right pocket.
TVWS in 2022
The people I really feel for in the TVWS world are the manufacturers of the access points and consumer terminals. Companies like Adaptrum, 6harmonics, Carlson Wireless and many others have gamely developed this innovative technology, trusting regulators to create the environment necessary for them to thrive. Many of them remain in business but struggle to reduce costs and invest in innovation due to limited demand.
In the United States, TVWS appears to be moribund. A report from the National Association of Broadcasters suggested, in 2021, that only 300 devices were connected to geo-location databases in the United States. To be fair, this is a claim from the broadcasters’ association so you have to take it with a grain of salt. But crucially, I haven’t seen this figure rebutted by anyone. In the UK, TVWS geo-location database service providers have apparently shrunk to just one, Red Technologies. There is no indication of how many devices are registered.
Here’s the current status of TVWS regulation in Sub-Saharan Africa:
|South Africa||Regulated||2018||South Africa has the CSIR to rely on for the geo-location database service. Officially, the CSIR is just supposed to be the reference database with a commercial geo-location database service provider expected to emerge but that has not transpired.|
|Mozambique||Regulated||2019||Mozambique experimented with a geo-location database service but realised that it was overkill both financially and administratively for their needs. Because so much television spectrum remains unoccupied in rural Mozambique, they were able to implement TVWS technology without a database, unless Excel counts as a database.|
|Kenya||Regulated||2021||Kenya passed TVWS regulations in 2021 but are still awaiting the final stamp of approval on a geo-location database provider.|
|Uganda||Stalled||2019||The Ugandan regulator (UCC) developed a draft standard for TVWS but nothing seems to have happened since then.|
|Nigeria||Stalled||2019||The Nigerian regulator (NCC) developed draft guidelines on TVWS but nothing seems to have happened since then.|
|Ghana||Stalled?||2019||The Ghanian regulator (NCA) developed draft guidelines on TVWS in 2019 but the document appears to have disappeared from their former link on the website. A local copy here for reference. Not entirely clear what the status is.|
|Malawi||Stalled||2015||From a pioneering role in TVWS trials in 2013, the Malawi regulator has not followed through in implementing TVWS regulation to date.|
|Tanzania||Stalled||2017||Another early adopter of TVWS trials in 2013, followed by an experimental license acquired by the University of Dodoma for TVWS in 2018 but no action has been taken by the regulator.|
|Morocco||Stalled||2015||Trials carried out in 2015 but no regulatory action.|
Looking through the list of countries and results, it is hard to escape the conclusion that the implementation of a geo-location database for TVWS is a significant stumbling block to the successful implementation of the technology. African regulators have struggled to contend with the imperfect TVWS regulatory model that emerged out of conditions unique to the United States, which raise the question as to whether there isn’t an opportunity for regulators in the region to define their own path to TVWS.
Don’t get me wrong, I think databases are the future of spectrum regulation. Well-designed databases are ideal for efficient management of scarce resources. However, there is an opportunity to act in parallel, to issue authorisations for TVWS spectrum in rural areas where television spectrum is dramatically underutilised, while simultaneously taking the necessary time and care to develop a sustainable geo-location database strategy.
How to Save TVWS
In spite of the perhaps bleak picture that I have painted of TVWS, I continue to think it is an important technology for affordable access to communication and that it is possible to design regulation that is fit for purpose. Here’s what I would do:
- Regulation Now, Database Later
Introduce TVWS regulation without the requirement of geo-location database operation initially. In most African countries, there is sufficient unused UHF spectrum, especially in rural areas, that manual assignments are possible. Whether that is managed internally via a spreadsheet or even a database, the point is to keep it in-house and simple while the market evolves. The exception to this rule is perhaps South Africa who already have a working database.
- Implement a temporary fee moratorium for ISPs
TVWS’s competition is cheap WiFi broadband. Until the market is established, ISPs should be exempted from any fees associated with TVWS in order to lower the barrier to adoption. The goal of the regulator should be to create a vibrant market for TVWS. Once the market is established, it is reasonable to review the fee structure.
- Simplify homologation and consider temporary import duty exemptions
Make TVWS manufacturers your friends. They have been running an ultra marathon waiting for regulation. Give them a sign of hope.
- Pursue Open Source for a shared geo-location database platform
If TVWS is to succeed, every regulator in the world faces the same challenge of affordably and securely implementing a geo-location database. This is the ideal scenario for an open source solution where development costs may be shared. It is also likely to improve security as many countries will be testing the limits of the same technology. Perhaps a philanthropy or development agency could be tempted to buy the geo-location software from the CSIR or Red or Fair Spectrum and release the code under an open license.
- Consider a regional approach
I have previously proposed a regional approach to geo-location database implementation. Regional regulatory bodies like CRASA, EACO, and WATRA are potentially well-positioned to support this. A regional approach would:
- significantly lower costs by sharing them across member countries;
- promote harmonisation of standards in TVWS regulation;
- allow the least resourced regulators to enjoy the same benefits as the best resourced; and,
- address any potential cross-border interference issues.
These actions may seem drastic to some and indeed one person has accused me of being a foreign agent for proposing dropping the initial requirement of a geo-location database but nothing could be further from the truth. I am proposing that African countries (and beyond) cease to feel constrained by hamstrung regulation that emerged out of a unique situation in the United States. I would like to add that while I have written about the regulation of TVWS in African countries (because that is the region I am familiar with) I think the same issues are likely to apply in much of Latin America and Asia.
I am indebted to my colleagues Adriana Labardini, Peter Bloom, Carlos Rey-Moreno, and Erick Huerta for feedback that improved this post. All errors remain mine. This work would not have been possible without the support of Mozilla and the Association for Progressive Communications. Views are my own.