Bringing affordable Internet to rural South Africa

To connect the rural unconnected to affordable internet is not so much a technical problem but more that of economics. For the large mobile operators, it is of no economic value to connect people who live away from the main rural highways in sparsely populated areas, even as a loss-leader project.

Technologically speaking it is not much of a challenge. Using TV white space (TVWS) in the 470-692 MHz frequency band on a secondary basis has proven an excellent technical solution, but the economics of the system are yet to be tested and that is what the new Wireless Access Providers Association (WAPA) project is about.

A consortium consisting of South African and US partners including Adaptrum, International Data Corporation (IDC), Microsoft, Project Isizwe and WAPA, together with the United States Trade Development Agency (USTDA) are working on a feasibility study. The project aims to demonstrate the technical, socio-economic, and commercial benefits of TVWS.

USTDA has provided $1-million to kick-start the project which was launched in Drummond in the heart of the Valley of a Thousand Hills in the KwaZulu-Natal Midlands, the most difficult terrain to provide reliable radio-based communication services. Consortium members are upbeat that the outcome will provide a workable solution and ready to scale to connect South Africans country-wide to the internet and to the educational, healthcare, and community economic benefits this brings.

WAPA feasibility project

The purpose of the initial feasibility study is to demonstrate TVWS as a commercially relevant and optimal solution for connecting rural South Africa. The end result of the feasibility study will be documented into a technical analysis, a financial feasibility and an executive white paper which will show:

  • The population density TVWS is best suited to serve.
  • The topology best-suited to using TVWS.
  • The average revenue per user which can be safely assumed once connectivity is in place.
  • The predicted return on investment periods.

WAPA project leader, Paul Colmer said that from this whitepaper, bankable projects across Southern and Eastern Africa can be developed. This will bring much-needed connectivity investment to the continent by tackling issues such as the cost of infrastructure needed to reach the rural poor.

The key advantage of TVWS deployment is that the signal coverage can reach up to 10 km radius from the base station without the requirement of line-of-sight. This makes it ideal for connecting people living in the rural undulating hills of KwaZulu-Natal and Eastern Cape, as well as areas in other provinces where there is poor existing backhaul infrastructure. Each TVWS base station will connect at a bandwidth of 20 Mbps to 30 WiFi access points. These will be scattered within walking distance of each citizen of the rural community. The project proposes deploying 1600 such base stations reaching 50 000 hot spots servicing a potential 13-million rural citizens.

Expected development impacts

The impact would be to provide at least 1 GB of data per person per month at a price of R30 per GB per month. The current cost of data in low income rural areas is about R600 per GB, so the new price would be less than 1/20th of the current price.

The expected coverage of the network will reach 13-million people currently living in 3,5-million dwellings in these areas. Thirty-five percent of these dwellings have access to the internet using a smart phone. The expected number of new subscribers is 1,7-million including some coverage in the hardest to reach, most dispersed, rural unconnected people from the Northern Cape.

One-year feasibility study

The initial one-year feasibility study will include a network build as well as the development of a business plan including financing aspects to help ISPs and their investors understand and take advantage of the commercial opportunity.

The second phase entails expanding the coverage of the commercial TVWS project into KwaZulu-Natal, Eastern Cape, Limpopo, Mpumalanga, Free State, North West and the Northern Cape by offering WAPA ISPs the opportunity to leverage the findings and other resources such as financing, developed during the first phase to scale out the solutions.

Multifaceted project consortium

WAPA is the South African WISP association representing 180 members. WAPA as well as Microsoft and Adaptrum are members of the Dynamic Spectrum Alliance (DSA), a global association that develops policy for and mechanism in bands such as TVWS. WAPA also has an MOU with the Department of Telecommunications, Postal Services and Communications to address affordable connectivity in the low-income regions.

Project Isizwe is a not for profit South African business. It has been operating for the past five years and has successfully managed free WiFi projects.

Adaptrum is a Silicon Valley based company with TVWS equipment, engineering training and installation expertise for the optimal deployment of the technical part of the project.

The International Data Corporation (IDC), a market research and analysis company, will provide an executive whitepaper and overall assessment of the technical and economic aspects of the TVWS solution.

Microsoft’s Airband team will provide the underlying support on TVWS policy as well as project management and technical architecture capabilities and services on cloud-based CRM and billing systems, on education solutions, as well as on the spectrum database.

The initial commercial feasibility study will use leased and loaned equipment from Adaptrum, Isizwe, and Microsoft. Three WAPA WISPs are supporting the project: Arband (part of Herotel). Letaba. and Mia Voice and Data.

The backstory of TVWS in SA

South Africa has been a driving force in TVWS from its early stages and is the first country in Africa to have published a TVWS regulatory framework. Encompassing an unreached population of over 20-million people in hard to reach terrain makes SA an ideal test location for a solution that scales affordably.

The September 2016 issue of IEEE Spectrum carried an article titled “Bridging Africa’s Broadband Divide”, in which authors David Johnson and Chomora Mikeka outlined trials in Cape Town and Malawi.

The trial in Cape Town broke new ground by using channels adjacent to those being used for live broadcasts, something that ran contrary to practice in most countries. Nevertheless, no reported interference to existing services resulted. The trial was mostly conducted by the Council for Scientific and Industrial Research (CSIR), using manual analysis as the Geolocation Spectrum Database (GLSD) was not ready at the time.

Since about 2010, the CSIR Meraka Institute had been working with ICASA, Sentech and others, punting the merits of white space systems. The Meraka Institute developed a sensing technology demonstrator to show the feasibility of the approach. In 2011, Google approached ICASA about implementing white space systems. ICASA introduced them to the Meraka Institute knowing that the required capability was available locally.

The Cape Town trial was the culmination of this process, in which the Meraka Institute cooperated with Google and other partners to develop and demonstrate the TVWS technology. The results of the trial were sufficiently successful that the trial was cited by the USA’s regulator, the FCC, as proof that adjacent-channel operations were feasible, contrary to what had previously been thought.

In 2014, Meraka Institute conducted another trial in cooperation with Microsoft, the Department of Science and Technology, the University of Limpopo and Multisource. Other African trials have since been run in Botswana, Ghana, Kenya, Malawi, Namibia and Tanzania.

A CSIR Meraka Institute team developed a Geolocation Spectrum Database (GLSD). Using this open standard and a comprehensive database of existing licensed transmitters, the GLSD is able to accurately forecast the presence of commercial signals at specific locations, thereby avoiding much of the potential for interference.