Unlocking the potential of unoccupied spectrum in developing countries: Southern African Development Community – case study

ABSTRACT

The lack of adequate telecommunications infrastructure in developing regions makes wireless systems the most feasible solution for providing Internet connectivity. With the global migration of television (TV) systems from analogue to digital, the TV spectrum is expected to be shared between broadcasting and broadband services. However, the absence of suitable regulatory policies for TV band sharing and low average revenue per user experienced by wireless network operators makes it difficult to deploy broadband networks in developing regions, especially in rural areas. This article presents a co-evolution analysis of three key domains of change (policy, technology and business) in providing broadband connectivity focusing on the Southern African Development Community region. Furthermore, the article proposes public–private partnership and public–private–people partnership models for deploying wireless broadband networks in developing regions. This article is useful to various stakeholders, including policy-makers, governments, the wireless communications industry and academia, by addressing the digital divide in developing regions.

KEYWORDS:

• Broadband access
• community networks
• public–private–people partnership
• Southern African Development Community
• television white spaces

1. Introduction

The digital divide between developed and developing regions is manifested not only by the delay in deployment of modern communication systems, such as wireless broadband and optical networks, but also in the lack of coordination between key policy-makers and the business sector. It has already been observed for many years that investments in telecommunication infrastructure for broadband access have significant impact on the economic growth and competitiveness of a nation (Czernich et al., 2011). Meanwhile, different scholars strongly agree that there is an urgent need for bridging the growing digital divide between developed and developing regions (Götz, 2013; Rajabiun & Middleton, 2013; Nucciarelli et al., 2013; Sridhar et al., 2013). This need is of highest importance in the Southern African Development Community (SADC) region where Internet penetration is as low as 7% (SADC, 2012) compared with over 75% in Europe (ITU, 2013). Providing cost-effective and reliable wireless broadband access to developing regions should therefore be a top priority for every government in the SADC region.

Rural areas are the most affected when it comes to the lack of broadband connectivity, and as such the use of wireless networks becomes a more cost-effective and practical solution than building new wire-line solutions (Masonta et al., 2012a). Thanks to the global migration of television (TV) systems from analogue to digital (a process known as digital switch over [DSO]), a considerable amount of the TV spectrum (or TV white spaces [TVWS]) will become available for providing broadband services on a secondary basis (ITU-WRC, 2012). Successful access or sharing of the TVWS spectrum depends on: reliable detection of white spaces; adequate protection of licensed TV band services or primary users (PUs) against any harmful interference; the amount of TVWS available in the area (before and after the DSO process); and sustainable business models which will attract investments in TVWS exploitation. For TVWS discovery and PU protection, cognitive radio techniques such as spectrum sensing and geo-location spectrum databases (GSDBs) are being considered, with GSDBs being preferred for the TV band (FCC, 2012; Ofcom, 2013; ETSI, 2013). Regarding the amount of TVWS availability, our previous measurements in both urban and rural areas in Southern Africa found that there to 97% of the TV spectrum is vacant (Masonta et al., 2012b; Lysko et al., 2012). What remains a challenge for exploiting TVWS is the absence of supporting regulations and policies from most regulators in Africa and other developing regions. On the other hand, necessary regulations can be developed once there are clearly defined and sustainable business models for exploiting TVWS to address the digital divide. Finding sustainable business models is especially necessary for funding large-scale deployment of information and communications technology (ICT) infrastructure to the target communities (Nucciarelli et al., 2010).

To address the high investment required to deploy broadband community networks, the public–private–people partnership (PPPP) model is proposed for community networks in developing regions. Unlike public–private partnership (PPP), the PPPP model ensures that community members or people are engaged in every phase of the project. This model has the potential of developing new skills and creating jobs for community members, some of whom can become village operators or value-added service providers. It has been reported that engaging local people in community projects increases the success rate of such projects and minimises property vandalism (Gumbo et al., 2012). The PPPP concept for wireless community networks was found to be successful in one of the oldest and most sustainable rural community networks in South Africa called the Siyakhula Living Lab (Dugmore, 2012).¹

This article contributes to the realisation of broadband access to developing regions with specific focus on the SADC region. Specifically, the article makes the following contributions:

• Providing an overview of the SADC region with specific focus on TVWS availability, regulatory hierarchy and DSO status.

• Analysing the potential of using the TVWS spectrum for delivering broadband Internet access in developing regions using the co-evolution process which takes into account three key domains of change in the communications sector: the regulatory policy, technology and business domains.

• Finally, the article presents PPP and PPPP models that can be adopted by developing regions, especially for TVWS network deployments in remote and rural areas.

The remainder of this article is arranged as follows. Section 2 provides a basic literature review on the use of TVWS for broadband access and an overview of the SADC region regarding the regulatory hierarchy, available TVWS and status of the DSO process. Section 3 presents an analysis of the technological domain, regulatory policy domain and business domain regarding the use of TVWS. Section 4 discusses PPP and PPPP models for fostering broadband access in rural areas. Section 5 concludes the article.

2. TVWS for broadband access and the SADC region: Overview

2.1. TVWS for broadband access

TVWS are portions of the Radio Frequency (RF) spectrum on the TV band that are not being utilised by licensed TV services as a result of frequency planning and as a by-product of the DSO process. According to the ITU-R 2012 World Radiocommunications Conference (WRC-2012) resolutions, the upper band of high frequency (UHF) TV spectrum (694 to 790 MHz) should be cleared as part of global DSO (ITU-WRC, 2012). In 2015, it was stated that this frequency band will be allocated to mobile services and was identified International Mobile Telecommunications (IMT) in Region 1; this also provides a harmonised worldwide allocation of this band (Maniewicz, 2016). In the same document, one may find that the discussions on possible regulatory actions in that band are considered in the agenda for WRC-23 for Region 1. As a result, TVWS are expected to be exploited on the lower UHF band (470 to 694 MHz). The use of TVWS for broadband communications has been promoted by leading ICT regulators in the USA and Europe over the past decade.

In the USA, for instance, the Federal Communications Commission (FCC) authorised licence-exempt access of TVWS for broadband connectivity, provided PU communication is not affected (FCC, 2012). The same approach was adopted by Ofcom in the United Kingdom (Ofcom, 2013). However, in the case of the TVWS spectrum, it is a dilemma for regulators in developing regions (such as the SADC) to decide which regulations should be adopted in their countries. The confusion or dilemma is due to the fact that most existing TV standards in the SADC region are aligned to European standards, while the TVWS regulations that can be palatable for developing regions are those proposed by the FCC. For instance, during the recent TVWS trials conducted in Cape Town, South Africa, there was some confusion on whether to adopt FCC or Ofcom rules/regulations when setting interference thresholds and other TVWS parameters (TENET, 2013). It is therefore important for developing regions to define clear policies and regulations in order to leverage the opportunity offered by the TVWS spectrum when addressing the broadband challenge.

In 2011, several spectrum measurements were conducted around urban and rural areas in two SADC member countries (Masonta et al., 2012b; Lysko et al., 2012). These studies found that about 70% and 50% of the TV spectrum was vacant in Pretoria and Cape Town, respectively. For rural areas, it was found that as much as 97% of the spectrum was vacant in Phillipstown (South Africa) and more than 97% of the spectrum was vacant in Macha (Zambia). Figure 1 shows a snapshot of the UHF frequency band in South Africa and Zambia. This profusion of TVWS in rural areas presents an opportunity for providing long-range broadband connectivity in rural areas, even with challenging terrain, such as mountains, vegetation and poor road infrastructure.

Figure 1. TV band scan showing spectrum occupancy in SADC urban and rural areas (Masonta et al., 2012b).

2.2. The SADC region overview

The SADC region is a regional economic community in Southern Africa formed in 1980, consisting of 15 member states.² Some of the challenges in the region include relatively high poverty, high rural populations and limited access to basic infrastructure and services, such as roads, electricity and ICT. However, in the decade 2001–10, six of the world’s ten fastest-growing economies (in terms of annual average gross domestic product [GDP]) were in sub-Saharan Africa, including two SADC countries: Angola at 11.1%, the highest in the world, and Mozambique at 7.9% (The Economist, 2011). The region achieved an annual average GDP per-capita growth of about 3% between 2003 and 2013.

Furthermore, the SADC region has a combined GDP of approximately US$470 billion and the Gross National Income per-capita values in the range of US$250 to 14 120. This article considers the SADC region because it represents a good example of a developing region with a population of about 298 million people, with an average 63% of rural population, and average Internet penetration of 7% (SADC, 2012). Table 1 summarises some key indicators for SADC countries based on data from the World Bank website.³

Table 1. Some key indicators for the SADC region.

Country	Population in 2014 (× 1000)	GDP per capita in 2014 (US$)	GNI per capita in 2014 (US$)	Rural population in 2014 (%)	Population with access to electricity (%) in 2012	Mobile subscription in 2014	Fixed broadband subscription in 2014	Internet users in 2014
						Measured per 100 people
Angola	24 227	5 900.5 (2013)	–	56.70	37	65.5	0.40	21.30
Botswana	2 219	7 123.3	7 240	42.80	53.2	167.3	1.60	18.50
DRC	74 877	442.3	380	58.00	16.4	53.5	–	3.00
Lesotho	1 897	1 034.2	1 330	73.20	20.6	85.0	0.10	11.00
Madagascar	23 571	449.2	440	65.50	15.4	41.2	0.10	3.70
Malawi	16 695	255	250	83.90	9.8	33.5	0.10	5.80
Mauritius	1 260	10 016.6	9 630	60.20	100	132.2	14.60	41.40
Mozambique	27 216	585.6	600	68.10	20.2	69.8	0.10	5.90
Namibia	2 402	5 408.2	5 630	54.30	47.3	113.8	1.80	14.80
Seychelles	91, 4	15 564.6	14 120	64.40	100	162.2	0.10	54.30
South Africa	54 001	6 483.9	6 800	35.70	85.4	149.2	3.20	49.00
Swaziland	1 269	3 477.1	3 550	78.70	42	72.3	0.40	27.10
Tanzania	51 822	955.1	920	69.10	15.3	62.8	0.20	4.90
Zambia	15 721	1 721.6	1 680	59.50	22.1	67.3	0.10	17.30
Zimbabwe	15 245	931.2	840	67.50	40.5	80.8	1.00	19.90

Notes: Data sourced from the World Bank website: http://data.worldbank.org/indicator/ Accessed 30 May 2016. DRC = Democratic Republic of Congo, GNI = Gross National Income

It can clearly be seen from Table 1 that huge gaps between SADC member states exist when it comes to some key indicators such as GDP, households with access to electricity as well as broadband penetration. These gaps are associated with the level of development of each member state. It is therefore important that, in order to address challenges facing the SADC region, member states should develop methods ensuring that each country within the region is supported to achieve higher levels of development.

2.3. Spectrum regulation hierarchy in the SADC region

Regional regulation in wireless communication is important to ensure collaboration and harmonisation, and to increase the economies of scale. Figure 2 depicts the regulatory hierarchy from the national level up to the United Nations level adopted by SADC member states. National regulators are tasked with regulating ICT and ensuring that the RF spectrum is allocated and managed efficiently in line with International Telecommunications Union outcomes. The Communications Regulators’ Association of Southern Africa (CRASA) is responsible for facilitating the development and harmonisation of ICT policies and regulations within the SADC region (SADC, 2012).⁴

Figure 2. Spectrum regulation hierarchy from national to ITU level (Masonta et al., 2013).

The African Telecommunications Union (ATU) is the continental organisation fostering the development of ICT infrastructure and services, with the membership of 46 countries and 17 associate members (ATU, 2012). All African countries are expected to make their submissions to the ITU through ATU. There are three ITU-R regions, and SADC falls under the ITU‐R Region 1 together with Europe, the Middle East, Russia and Mongolia. For the entire world, RF spectrum regulation decisions are taken at the ITU-World level through the ITU-R World Radiocommunications Conference meetings which are held every three to four years in Switzerland.

3. Co-evolution analysis

Over the past two decades, the co-evolution process has been widely used to evaluate the introduction of new technologies, such as spectrum sharing in mobile networks (Ahokangas et al., 2013; Lyytinen & King, 2002), and cognitive radio systems (Fomin et al., 2011). Co-evolution refers to a large systemic innovation that demands the coordination of independent and heterogeneous players to ensure compatibility and interoperability across different systems (Lyytinen & King, 2002). The three key domains of change in the wireless broadband communication environment are regulatory policies, technology and business, as shown in Figure 3. According to the co-evolution process, at least two or three domains are said to co-evolve as long as each has a significant impact on the other. This article uses the co-evolution process to analyse regulatory policies, technology and business as the key three domains of change in the wireless broadband communication environment.

Figure 3. Co-evolution dynamics for the TVWS access ecosystem.

In our case, the policy domain includes government ministries of ICT and spectrum regulatory agencies, the technology domain includes standardisation bodies, cognitive radio systems and white space devices (WSDs), and finally the business domain includes national or local wireless Internet service providers and PPP/PPPP models. As will become clear in the following sections, the three domains are overlapping by nature. For instance, GSDBs overlap all three domains. As a result, a unified vision is required across all three domains; hence we propose PPP and PPPP models because of the interaction of the domains within the TVWS eco-system.

3.1. Technology domain analysis

There are ongoing standardisation efforts focusing on the use of TVWS for broadband technology. Some of these are suitable for urban areas, while others suit rural and remote areas. This section analyses two Institute of Electrical & Electronics Engineers (IEEE) standards or technologies which are potential candidates for TVWS network deployments in developing regions.

3.1.1. IEEE family of TVWS standards

IEEE 802.22 is the wireless air interface standard focused on the development of cognitive radio-based wireless radio access network physical and medium access control layers for operation in TVWS (IEEE 802.22, 2011). This standard specifies a point-to-multipoint wireless air interface where a base station manages all associated customer premises equipment. Current investigations within the group concentrate on the definition of the Standard for Spectrum Characterization and Occupancy Sensing, and are conducted under the acronym 802.22.3 (IEEE P802.22.3, 2014). A typical use case for 802.22 would be in sparsely populated rural areas, which makes it a suitable candidate for providing rural wireless broadband access in the SADC region.

IEEE 802.11af is another possible and promising standard for TVWS access. IEEE 802.11af is a modification to the 802.11 physical and medium access controls for operation in TVWS (IEEE 802.11af-2013, 2013). A recent large-scale study on the feasibility of 802.11af shows that its operation in rural areas is viable, and can provide broadband at speeds of over 54 Mbit/s (Latkoski et al., 2012). Both IEEE 802.22 and IEEE 802.11af use the GSDB technique for TV incumbent protection, as well as for TVWS channel discovery. A typical deployment of IEEE TVWS standards in a rural area is shown in Figure 4 with backhaul options.

Figure 4. Proposed rural TVWS network with three options for providing Internet connectivity to the TVWS base station to access the GSDB. Note: CPE = customer premises equipment.

The IEEE 1900.x series standards are managed by the IEEE DySPAN-SC with specific focus on dynamic spectrum access topics. The scope of IEEE DySPAN-SC includes the coordination of wireless technologies, including network management and information sharing amongst networks deploying different wireless technologies (DySPAN-SC, 2014). One of the relevant standards in the 1900.x series is IEEE 1900.7 ‘for radio interface for white space dynamic spectrum access radio systems supporting fixed & mobile operation’ (DySPAN-SC, 2014). Current investigations focus on the ‘Standard for Spectrum Sensing Interfaces and Data Structures for Dynamic Spectrum Access and other Advanced Radio Communication Systems’.

3.1.2. Backhaul solutions

Most of the areas which lack Internet connectivity are situated in remote and isolated locations far away from the cities. Thus, one question commonly asked when bringing wireless connectivity to such areas concerns the best backhaul technology to be applied. There are several technologies that can be used to address this question. This includes the use of TVWS as backhaul, fibre optic, point-to-point microwave links, satellite, copper solutions and free space optics. In some cases, the use of existing cellular infrastructure can be used by WSDs for reaching the GSDB, while other options include point-to-point microwave links, satellite links or existing national or private optic fibre running through some villages. These technologies are summarised in Table 2.

Table 2. Common backhaul technologies for wireless networks.

Technology	Link distance (km)	Capacity	Backhaul rollout experience		Characteristics
			RF licence	Time to deploy	Advantages	Disadvantages
Point-to-point microwave links^a	∼50	>1 Gbit/s	Both licence required and licence exempt	Fast rollout	Mature technology; widely used for cellular backhaul; high throughput; rapid deployment	Requires Line-of-sight and high site mast/structure; requires RF licence
Satellite link (VSAT)	From satellite	∼4 Mbits/s (downlink)	Licence required	Fast rollout^b	Quick to set up; reliable connection; mature technology	Expensive monthly rental fee; limited throughput (low)
IEEE 802.22 TVWS link	∼100	22.69 Mbit/s	Licence exempt	Fast rollout	Uses TVWS – great propagation; no spectrum licence required	New technology; throughput limitations; new high sites for base station installation
Fibre optic^c	>100	∼1 Pbit/s	Not required	Slow rollout	Most reliable; high broadband; mature technology; immune to cable theft	Very expensive; takes long time to implement; way-leave required to dig trenches or plat poles
Free space optics	<10	∼10 Gbit/s (depends on distance)^b	Not required	Fast rollout	Quick to set up; no licence required	New technology; sensitive to weather (e.g. rain and fog); short distance range; requires line-of-sight (LOS)
Copper/coaxial cable-based technology^b	<3	60 Mbit/s	Not required	Slow rollout	Suitable for metropolitan areas; reliable connection	Cable infrastructure not available in most rural areas; expensive to rollout; cable theft risk; required shorter distance to the exchange

^aSource: http://www.digitalairwireless.com/files/Fiber-vs-Microwave-White-Paper_1333235596.pdf.

^bSource: http://www.brocade.com/downloads/documents/technical_briefs/next_gen_wireless_backhaul_ga_tb_188_00.pdf.

^cSource: http://www.ntt.co.jp/news2012/1209e/120920a.html.

3.1.3. Summary of technology domain

In order for the SADC region to benefit from economies of scale, it is important for member states to coordinate their choice of standard technology. Such an approach will also support a seamless handover of end user devices across borders. Although most regulators have adopted a technology-neutral approach to issuing licences, there is a need to coordinate decision-making for brand new technologies, such as TVWS systems. The CRASA should play a role in providing advice to member states’ regulators, as well as in the coordination of technology policy decisions.

3.2. Regulatory policy domain

There are currently two mechanisms being considered for PU protection: GSDBs and spectrum sensing using cognitive radios (Mwangoka et al., 2013; Fitch et al., 2011). Solutions for fine spectrum sensing algorithms are still a subject of ongoing research, whereas the GSDB mechanism is more established and proven, especially for TVWS. A combination of GSDB and spectrum sensing remains a promising approach to regulating opportunistic TVWS access. A comparative study on existing TVWS regulations adopted by the FCC and Ofcom was conducted in our prior work and published in Masonta et al. (2013). This subsection analyses mechanisms that can be used by national regulators for protecting licensed services in the TV spectrum band.

3.2.1. Protection of digital terrestrial TV signals

Protection of TV broadcasting against harmful interference that might be caused by WSDs is among regulators’ top priorities. This includes both analogue and digital terrestrial TV (DTT) broadcasting. Currently, the FCC and Ofcom prefer making the use of GSDB mandatory for anyone intending to exploit TVWS for broadband services. Currently, there are no policies governing the exploitation of TVWS in the SADC or most developing countries.

3.2.2. Protection of programme-making and social event devices

The use of GSDBs is suitable for TV services, because the power of the broadcasted signal is high and the TV infrastructure is stable. An immediate problem appears when other co-existing services are taken into account. In particular, the use of WSDs in the presence of active programme-making and social event (PMSE) devices, such as wireless microphones or cameras used for event recording and transmission, is a big challenge. Unlike static TV infrastructure, PMSE devices are not bound to one specific location.

In this context, two paths which SADC regulators can follow are considered, as proposed in Masonta et al. (2013). The first concept assumes that dedicated (or safe-harbour) TV channels are reserved only for PMSE and radio astronomy services, minimising the amount of interference collected from neighbouring signals. However, all active PMSE devices operate in the whole TV band, because ensuring backward compatibility would be impossible if only one TV channel was allowed for such transmission. A transition period would be required. Another solution focuses on the existence of a dedicated PMSE database. If the usage of particular devices can be stored in the PMSE database, the quality of signal transmission can be guaranteed; that is, white space users will get information about the presence of PMSE devices in their vicinity. Conversely, the lack of information in the PMSE database will result in a situation where PMSE transmissions could be disturbed by secondary services.

3.2.3. Protection of other services operating inside or close to the TV band

Besides DTT services and PMSE transmissions, there are other wireless systems operating in the TV band, 470 to 790 MHz, such as radio astronomy. Because the SADC falls within the ITU-R Region 1 together with the European Union, this subsection provides general guidelines for consideration in protecting other TV band services and adjacent services. Following the analysis from the European Communication Commission, it can be foreseen that very large separation distances between the WSDs and radio astronomy would be required if operating in TV channels 37 to 39 (CEPT, 2011). The European Communication Commission recommends that TV channels 37 and 39 should be excluded from the usage by autonomous WSDs based on spectrum sensing only. The usage of databases is highlighted. When adopting this rule in the SADC region, one can conclude that it would be beneficial to exclude channel 38 from the autonomous WSD transmission, and implement databases for the neighbouring channels 37 and 39. Recently, Ofcom released a document that enables ‘the use of certain wireless telegraphy equipment complying with the technical parameters set out in the WSD Regulations on a licence exempt basis’ (Ofcom, 2015). The consultation process on this subject was still open while preparing this article.

3.2.4. GSDB as a spectrum management tool

Regulators are expected to ensure that GSDB owners comply with applicable regulations, the data provided by GSDBs are accurate and the GSDBs are always available. One of the possible solutions, concentrating on the licence-exempt TVWS access, is to implement the vision adopted by Ofcom (CEPT, 2011). In such an approach, the WSDs belong in one of two categories: master devices and slave devices. Master devices must have a reliable Internet connection to obtain the list of qualified GSDBs and the link to query selected qualified GSDBs.

A slave device, on the other hand, does not have access to GSDBs, and receives all necessary parameters for transmission through the master device. This scenario is illustrated in Figure 5, and is proposed for the SADC region.

Figure 5. Framework for authorising the use of TVWS by the regulator based on the list of registered and qualifying GSDBs, as adopted from Ofcom (2013).

3.2.5. Summary of regulatory policy domain

This subsection summarises how policy can be used to address some of the major challenges faced by SADC member states as follows:

• Creation of policies to allow automated cross-border RF spectrum coordination techniques to be developed. This will ensure that there is RF spectrum harmonisation which will lead to efficient utilisation of the RF spectrum among neighbouring countries. Such policies should be implemented by CRASA which can make use of systems such as GSDBs.

• Lack of policies guiding the use of TVWS can be seen as one of the major issues affecting the deployment of TVWS networks and technology within the SADC region. Although countries like South Africa and Malawi have already started initiatives to develop such regulations, it would be beneficial if such regulations were also coordinated at CRASA level to ensure harmonisation.

• Protection of other primary services against potential interference from the TVWS network should also be managed at a regional level, given the greater propagation properties of TVWS. Again, the responsibility to manage such protection of TV band primary services will be more efficient if coordinated by CRASA.

3.3. Business domain analysis

The business domain in the TVWS ecosystem for the SADC region shall revolve around spectrum markets, as well as manufacturing of WSDs and intelligent set-top boxes (STBs).

3.3.1. GSDB ownership

From the technical viewpoint, a GSDB contains records of the protected contours of PUs. From the management perspective, we propose that ownership of GSDBs should be in the hands of third-party companies, which is the model adopted also by the FCC (2008), and considered by Ofcom (2015). Allowing third parties to have ownership of GSDBs will make the market more competitive and brings more innovation, as opposed to a situation where such ownership is given to the government or regulators. This means that a country will be filled with several GSDB providers. Therefore, it becomes important to ensure interoperability of these GSDBs, so that any WSD can be able to query any local GSDB for spectrum availability. One of the solutions is the usage of the so-called spectrum brokers. In view of the recent TVWS trials performed in South Africa (TVWS Trials, 2013), Malawi, Tanzania and Kenya, multinational technology organisations (e.g. Google and Microsoft) can be viewed as potential players in the GSDB ownership/hosting market. However, SADC governments can also promote the ownership of national and regional GSDBs to be in the hands of local entrepreneurs or businesses. This will help SADC member states to address poverty by creating jobs and encourage the participation of local entrepreneurs (Kliks et al., 2016). Various concepts on how the databases could be used for spectrum management were presented in papers by other authors, such as Perez-Romero et al. (2015).

3.3.2. Secondary spectrum access through spectrum brokers

A spectrum broker’s main role is to control the amount of spectrum bandwidth and transmission power assigned to each WSD in order to keep the desired quality of service above and interference below the regulatory limits. A spectrum broker can act as an intermediary between a GSDB and players that negotiate the spectrum on behalf of spectrum users among different areas in the SADC region. This approach was first proposed in Europe through the COGEU (COGnitive radio systems for efficient sharing of TVWSs in EUropean context) project (Lavaux, 2011; Mwangoka et al., 2011). Basically, a secondary spectrum market is managed in the form of auctions by a dedicated legal body or spectrum broker. Figure 6 depicts the COGEU approach to spectrum brokers.

Figure 6. Spectrum broker concept based on the COGEU project (Lavaux et al., 2011).

3.3.3. Local STB manufacturing

Local manufacturing and distribution of STBs is another new business opportunity that could be realised by opening up TVWS in developing regions. STBs are important for the conversion of DTT signals to analogue signals that can be viewed on old analogue TV sets. In order to understand the potential market size of STB distribution in developing regions, the current statistics on TV subscribers is analysed using South Africa as one example.

Prior to the completion of the DSO process, two methods are commonly used to receive TV services in South Africa: free-to-air terrestrial analogue TV and paid digital satellite TV (Armstrong & Collins, 2004). About 75% of the estimated 11.5 million TV-owning households received free-to-air analogue TV services (DoC, 2012). The migration of the 75% free-to-air analogue TV subscribers to DTT requires an upgrade of analogue TV sets to DTT sets. In 2012, the South African DoC conducted a study which revealed that over 5 million TV-owning households were poor and could not afford to buy the STBs (DoC, 2012). Furthermore, the study discovered that the majority of these poor households are in rural areas. To facilitate smooth digital migration, the government decided to subsidise these poor households.

3.3.4. Summary of the business domain

From the SADC region point of view, it would beneficial if the opportunity presented by the TVWS technology could be harnessed by creating new business models that will provide services across borders with less red tape. This will contribute to reducing the cost of doing business in the region, while at the same time reducing unemployment rates. Again, in order to increase local skills and business, SADC member states are expected to subsidise local entrepreneurs who will be manufacturing smart STBs (SADC, 2010).

Figure 7. Challenges facing SADC countries in deploying TVWS networks.

4. Proposed PPPP model for developing regions

Based on the above discussion, we can identify a set of key challenges faced by developing regions. These can be briefly classified as shown in Figure 7 and as follows, focusing on the SADC region:

• Lack of infrastructure: the biggest problem lies in the lack of infrastructure and poverty in the region; there is thus a need for coordinated actions and cooperation between all—government, industry and people.

• Technology: there are several technological solutions available all over the world, and a decision should be made on the selection of the best option for the given region.

• Policy: the decisions mentioned should be made in consultation with key industry players, but ultimately such regulations are the responsibility of governments and ministries, taking into account relations with neighbouring countries and international agreements. Furthermore, appropriate political mechanisms have to be offered to industry and communities to accelerate the adoption of the selected technology.

• Industry/business: typically, industry players lobby for certain technologies; the coordination between policy-makers and industry players is thus important. On the contrary, to ease the implementation of certain technological solutions, an appropriate legal framework should be implemented to support the desired development direction (e.g. creating economic zones for small–medium enterprises).

• End-users: finally, end-users have to be included, at least indirectly, in the entire process, because satisfying end-users’ needs will be the final criterion for the success of the whole venture. It is also the community that should be motivated and inspired for local enterprise.

Thus, in this article we propose to consider PPP and expand our version of the PPPP model (as proposed in our early work [Ramoroka et al., 2016]) as the solution for developing countries. Acknowledging that several solutions might exist for addressing the broadband challenge, this article focuses on the development of solutions based on partnerships between the public sector, the private sector and the community or people. PPPs are considered to be the best option to foster development where there are insufficient investments and growing pressures on government budgets for service provisions (African Development Bank, 2004). PPPs are used to strengthen and fasten the cooperative ventures between the public and private sectors. What makes PPPs more appealing are their expected benefits, which include greater value for money in public projects, clearer project objectives, improved incentives for competitive tendering and access to private finance for expanding services (Jamali, 2004).

On the other hand, the PPPP model is built on the living lab methodology where multiple stakeholders, including government, industry, research institutions and communities, collaborate in different stages of research, development and innovation (Gumbo et al., 2012). A PPPP is then formed to achieve mass impact when addressing societal challenges.

In order to address Internet connections in rural and remote areas in developing regions, a blend of PPPP and PPP models is encouraged. For instance, providing Internet connection to schools, health centres, rural government centres and rural business centres might require a formal agreement which can be achieved through the PPP model. On the other hand, the PPPP model might be suitable for connecting households within communities.

4.1. Generic PPP model for broadband access

The past two decades have witnessed a growing number of implementations of the PPP concept by municipalities in developed and developing counties for fostering broadband access to communities. Various PPP models can be identified depending on the degree of private sector involvement and risk; for example, the classification of PPP models used in Canada can be found in Canadian-Council-for-Public-Private-Partnerships (2007), while models used in the USA with the purpose of emergency management in Federal Emergency Management Agency (2013). Comprehensive guidelines on the PPP concept are presented in ESCAP (2011). Of highest relevance to our article is the work by which examines the emerging models for PPP in providing broadband access to communities (Nucciarelli et al., 2010). The following four distinctive phases of the PPP model are proposed:

• Phase 1: identification of social and economic targets by the municipality. The aim of this phase is to identify shared and competing goals by stakeholders, and match them with crucial policy issues (Mandviwalla et al., 2008). It is in this phase that stakeholders, especially public and private, take into consideration the potential demand and supply for services to be offered by the planned network.

• Phase 2: identification and matching of core resources and competences. The focus of this phase is to identify possible local community entrepreneurs or champions who will be involved in the project. The key outcome of this phase is a public–private collaboration which is used to transform the strategic vision into cooperation where the management of funds and risks is shared.

• Phase 3: TVWS network deployment. This phase considers stakeholders at different levels, and captures their vision, mission and competences to deploy the network. Depending on the choice of the PPP model adopted, each stakeholder’s involvement will differ based on their contributions to the funding, the adopted business model and the choice of technology used to build the network.

• Phase 4: provision of broadband services. Multiple broadband services and applications are provided to users. During this phase, local community members and entrepreneurs are encouraged to develop local content to share or upload to the network.

In order to start the next phase, the current one should be completed or in an advanced stage. Moreover, there is a feedback loop from the last phase to the initial one, allowing for updates of earlier decisions. Table 3 presents a summary of the status regarding each PPPP phase within the SADC region in the context of TVWS network deployment.

Table 3. Status and responsibilities for realising phases of PPP model.

PPP/PPPP phase	Actions taken or required	Responsibility	Status in the SADC
1: Identification of social targets for TVWS network	In 2012 the SADC region published the ‘Digital SADC 2027’ which considered TVWS for broadband access (SADC, 2012) All SADC member states realise the need to use TVWS to increase Internet penetration in rural areas Since 2012, several SADC countries participated in TVWS trials (such as South Africa, Tanzania, Botswana and Malawi)	The public sector in consultation with the people It is also important to engage the private sector at this early stage	In an advanced state
2: Identification and matching of core resources and competences	There is a need for a common approach to access TVWS Many SADC member states focused on completing the DSO process TVWS and broadband are becoming part of national agenda items There is a need for acquiring core competences especially for developing and managing PPP models	Executed by the public sector through engagement with the private sector and the people (community) Some useful resources may be sourced from the community	At the early stages in most member states
3: TVWS network deployment	Few TVWS network trials are deployed on a small scale (e.g. Cape Town TVWS trial) Large-scale deployment expected post 2015	Executed by the private sector with support from the public sector and the people	Still at the trial stages in a few SADC member states
4: Provisioning of broadband services	To be realised once Phases 2–3 are completed	Executed by people (community) with support from the private and public sectors	Expected after 2015

4.2. PPPP model for community networks

In order to achieve the ‘Digital SADC 2027’, all governments within the SADC region are expected to partner with the private sector and co-invest in rolling out broadband networks in rural and remote areas, whereas the private sector will solely invest in the urban areas. Such co-investments will encourage the private sector, especially small and medium enterprises and village entrepreneurs (known as village operators), to enter the ICT market by providing connectivity to small areas. We believe that the PPPP model is suitable for connecting households within the rural community. Furthermore, community networks are mainly made for sharing information among community members, and the network can be easily expanded as the community grows. Details of the PPPP model are presented in our previous work in Ramoroka et al. (2016).

4.3. PPP model for key community centres

In order to connect schools, government centres, business centres and health centres in rural areas, an experienced operator is required. This is mainly due to the need for reliable and secure connectivity, especially for health services and e-government where life-critical, personal and sensitive information is exchanged. Given that the current cellular infrastructure covers the majority of the population in many developing countries, national cellular operators can use their existing base stations to deploy TVWS links that connect rural schools, health centres or clinics and government centres.

In this case, the PPP model becomes suitable, whereby municipalities or government departments can partner with the private sector by developing incentive mechanisms to encourage national network operators to build and operate rural networks. For instance, governments can be responsible for all radio equipment at schools, health and government centres, whereas network operators will be responsible for providing wireless links and backhaul using TVWS on a licence-exempt basis. Because spectrum licence fees are important costs in providing wireless services, licence-fee exemptions for TVWS links for rural areas and communities will encourage national operators to provide rural broadband connectivity. The government can also introduce tax breaks for all national operators involved in rural broadband connectivity initiatives. For TVWS links provided by national network operators, there might be a need for coordination between local GSDB providers and spectrum brokers, and national GSDB and spectrum brokers.

5. Conclusion

This article studies the possibilities of unlocking the potential of the unoccupied TVWS spectrum for providing broadband access in developing regions, with a specific focus on the SADC region. The article starts by presenting an overview of the SADC region based on the availability of TVWS spectrum, the radio spectrum regulatory hierarchy and the status of the DSO process. The article then uses the co-evolution process to analyse the three key domains of change in the ICT: regulatory policy, business and technology. Finally, PPP and PPPP models for deploying TVWS networks in the rural areas of developing regions are discussed and analysed. To address the digital divide, developing regions should have a holistic approach that addresses key issues within the policy, technology and business domains. Because of lack of technical skills in rural areas, the establishment of living labs, based on PPPP and PPP models, can be a vital enabler of rural broadband connectivity for developing regions.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This is work was supported by the Polish Ministry of Science and Higher Education [grant number DSPB-08/81/158, DSPB-08/81/164]; South African Agency for Science and Technology Advancement [grant number UID: 84201, UID: 86432].

Notes

1 See www.siyakhulall.org

2 Angola, Botswana, Democratic Republic of Congo, Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Seychelles, South Africa, Swaziland, Tanzania, Zambia and Zimbabwe.

3 See http://data.worldbank.org/indicator/

4 CRASA represents all of the national regulators in the SADC except Madagascar and Seychelles.