Ten Years of TV White Space Advocacy

On Friday, March 23rd 2018, the South African communications regulator (ICASA) formally gazetted Regulations on the Use of Television White Spaces spectrum.  This means that, subject to type approval and to authorisation through a geo-location database, TV White Space (TVWS) wireless communication equipment can legally be used in South Africa.  This is a pretty big deal.  TVWS regulation exists in the United States, the UK, and in Singapore but South Africa is arguably the first market where TVWS technology could make a real difference thanks to a combination of need for affordable rural access to broadband and the relative abundance of unused television spectrum in rural areas in South Africa.  It is the first country in Africa to enable these regulations pushing ahead of Ghana and Malawi, both of which have draft TVWS regulations under review.  Hopefully other African countries will now follow suit.
Like WiFi, TVWS technology now has a very low barrier to adoption in South Africa.  It is not as permission-less as WiFi in that TVWS must connect to a geo-location database in order to check on what spectrum is available in the area that it is operating but still represents a technology that now any wireless ISP can adopt.  The potential of TVWS lies in the spectrum band it occupies, from 470MHz to 690MHz.  Unlike WiFi these frequencies do not require direct line-of-sight to deploy.  They can penetrate foliage and even transmit over hills.  This can mean significantly lower costs for towers that don’t have to be built as high.  TVWS is not a panacea though; the lower frequencies require larger antennas and cannot transmit the same capacity of broadband that something like 5GHz WiFi can but it represents an important complementary tool for solving affordable connectivity challenges.
More importantly, it represents the tip of the iceberg of a regulatory change towards using software to manage spectrum assignment dynamically.  If you think about how Airbnb has changed the hotel industry or Uber the taxi industry, the same potential exists to apply resource management software and business models to the process of spectrum assignments.  Part of the reason that Uber was so disruptive of the taxi industry is that it was so resistant to change.  The cost of a taxi medallion, which is simply a permit to operate a taxi in New York City, peaked in 2013 at 1.3 million dollars.  It was a market that was artificially constricted by the Taxi and Limousine Commission.  The current price is close to a tenth of that now.  There are many things that software is not so good at but dynamic resource allocation is not one of them.  Come soon or come late software is going to change the way we regulate spectrum. Through mechanisms like TVWS regulation, communication regulators have an opportunity to manage change for lower risk and less disruption.   South Africa is now ahead of the curve.
You may be wondering why I would describe TVWS as less risky since it is a new and relatively untested technology at scale.  This is because TVWS is regulated for use on a secondary basis.  That means that if a television broadcaster wants to use the frequency, it is theirs to use and the TVWS device must relocate to an unused frequency, as determined by a geo-location database.  National, exclusive-use, spectrum licenses which typically auction for millions of dollars and last for ten or fifteen years, are extremely high-risk for regulators.  What if the winner of the auction doesn’t live up to their promise?  Getting spectrum back and re-auctioning is a long and sometimes impossible process for regulators.  The regulator effectively shares the risk with the operator.  In contrast, TVWS regulation places all the risk where it should be, on the market.  If TVWS doesn’t turn out to be the amazing technology that we all hope it will be, the downside for the regulator is comparatively small.  This is not to say that TVWS technology is not risky in itself.  Any new technology is high risk until it matures in the market.  The key is who bears the risk.
As I was reading the news of ICASA’s decision, it occurred to me that I have been lobbying for TVWS regulation for just a few months shy of ten years. I don’t pretend to be anything but a small cog in the process of change but here are some things I remember. As a fellow at the Shuttleworth Foundation in 2008, I wrote my first article on TVWS in September 2008 after reading about Sascha Meinrath’s advocacy for TVWS in the United States.  I found common cause with the Wireless Access Providers Association (WAPA) whose members had good reason to desire more affordable access to spectrum.  The following year I wrote a cheeky series of articles entitled “What Google Should in Africa” which came out of my frustration with what seemed like a lack of strategic direction in Africa on the part of the Internet giant.   The first article in that series was a plea to Google to take up the same kind of advocacy for TVWS in Africa that they were doing in the United States.  I continued writing about its potential on the continent. The local press began to take notice.  Brainstorm Magazine interviewed me on the topic in 2009 but at that time TVWS must have sounded a little far-fetched.  And there, despite many discussions things lay.  In September 2010, the United States enacted TVWS regulation but the US seemed like a long way away in terms of policy and regulation. I continued to write and talk about the potential of TVWS but it wasn’t until early 2011, when Ory Okolloh, who had just been hired to be Google’s Policy Manager for Africa, reached out to me. It turns out that my series of cheeky blog posts were not completely un-read.  Ory and I met in Johannesburg and we discussed what might be feasible. In October 2011, together with the Association for Progressive Communications (APC) and WAPA, we organised a workshop to bring international experts, regulators from the region, and South African government, industry and regulatory folk to discuss TVWS and its potential.
Following the success of the workshop, a technology trial seemed like the logical next step to prove to ICASA that the technology would not cause harmful interference.  Ory lobbied Google internally to find financial support for a trial while I looked for a likely partner to host such a trial and a potential beneficiaries of the connectivity.  I approached Duncan Martin, CEO of TENET (Tertiary Education and Research Network of South Africa) which operates the university Internet backbone and he agreed that TENET might both host the trial as well as provide backhaul connectivity for it.  TENET was already partnered with the e-Schools Network whose mission is to provide access to secondary schools.  Providing connectivity to secondary schools seems like a practical choice for the trial.  It also seemed natural to involve the Council for Scientific and Industrial Research (CSIR) who were already doing research in this area to carry out rigorous spectrum interference measurement to ensure that the results of the trial would provide the most credible possible evidence. Critically CSIR was also a state agency which would carry weight in both the policy and regulatory spheres.  WAPA was involved on the advocacy side and communication service provider, Comsol, implemented the trial.  I think Google spent about $400K in total on the trial and just as important as the money itself was the attention that the Google brand brought to the issue.  This trial would never have happened were it not for the commitment and good will of all of the partners.  It took nearly a year to pull together but the trial began in September 2012 with ICASA granting permission to CSIR and partners to establish a TVWS network in the Western Cape.  It ran for just over a year and was a success in almost every respect. TENET has done a great job of documenting the trial.
Meanwhile in Kenya in 2013, Microsoft was busy sponsoring a TV White Space pilot of their own in Nanyujki.  Mawingu had the full support of Bitange Ndemo, the Permanent Secretary in the Ministry of ICTs and was the first of many TVWS trials that Microsoft has sponsored across the continent.  It seemed an opportunity to make common cause and both Microsoft and Google got together to sponsor an event in Senegal to raise awareness of TVWS technology and its potential for improving affordable access.  At that time, both TVWS technology and the associated empowering regulation seemed inevitable.  Having two rival technology titans like Microsoft and Google on the same stage making common cause was too good to be true.  Seizing this opportunity, Microsoft catalysed the establishment of a industry advocacy association for dynamic spectrum called the Dynamic Spectrum Alliance.  Dynamic spectrum is the more general term for TVWS, which refers to the specific application of dynamic spectrum management to television broadcast frequencies.  The Alliance went on to organise a series of successful annual global events on dynamic spectrum as well as drafting model regulations for regulators to examine and build on.
And yet change was slow in coming.  In Kenya, the communication regulator (CCK) was not nearly as enthusiastic about TVWS as the Ministry was and proved reluctant to allow use of television spectrum to Mawingu. When Mawingu’s experimental license expired, the regulator declined to renew it, forcing them to revert to a WiFi-only strategy.  To Mawingu’s credit they are going from strength to strength on WiFi alone.  The ITU took a very conservative perspective on TVWS, urging regulators to join ITU working groups on the topics and to wait until best practice had been established.  ITU’s message of fear, uncertainty, and doubt regarding TVWS inevitably had a chilling effect on progressive regulators wishing to move forward.  The industry association for mobile operators (GSMA) also joined the ITU in their dismissiveness of TVWS. The net result was delay.  Many of us thought that 2013 would be the year for TVWS to take off and then 2014, and then 2015. By 2016 I began to wonder what it would take for TVWS regulation to happen.
By this point, Google had lost interest in TVWS and shifted their attention to the 3.5GHz frequency band for dynamic spectrum.  Facebook briefly dallied with TVWS but declined to get involved.  Microsoft to their enormous credit, doubled down on their commitment and supported even more trials across the continent.  Their support of this technology has been critical for technology manufacturers, advocates of regulatory reform, and for regulators.  The Dynamic Spectrum Alliance (DSA), which Microsoft deserves the credit for both its initiation and ongoing support, has been particularly important in changing hearts and minds on TVWS.  But still I think by 2017 we all began to wonder whether TVWS was going to happen or not.  The DSA shifted its advocacy to be more inclusive of other approaches like unlicensed (WiFi) spectrum which was spreading rapidly in African countries.
And indeed the jury is still out but this decision by ICASA is a landmark. It is risky being first and by leading the way in Africa, South Africa has made it easier for regulators in other countries in the region to move ahead.  Hopefully this is also a signal to manufacturers that there is a market to manufacture for and that there will probably be many more to come on the continent.  It is a decision that fills me with hope and optimism.
Disclaimer:  The above is inevitably a very one-sided tale. Success has many parents and I am more like a noisy uncle in this context.  I wrote this partly to document my own history but mostly to remind myself that change that seems impossible and unlikely can happen but it is often a long game.  Here’s to persistence!

Image courtesy Dan DeCook