TV white spaces exploitation through a bicameral geo-location database

Telecommunications Policy 37 (2013) 116–129

Contents lists available at SciVerse ScienceDirect

Telecommunications Policy URL: www.elsevier.com/locate/telpol

TV white spaces exploitation through a bicameral geo-location database Joseph W. Mwangoka n, Paulo Marques, Jonathan Rodriguez ~ ´rio de Santiago, 3810-193 Aveiro, Portugal Instituto de Telecomunicac- oes, Campus Universita

a r t i c l e in f o

abstract

Available online 9 October 2012

The exploitation of TV white spaces can meet the increasing demand for spectrum resources and create opportunities for deploying a variety of wireless services in a flexible manner. However, uncertainties from technologies, business models and regulatory policies hinder the take-off of TV white spaces exploitation. This paper proposes a bicameral (or two-chambered) geo-location database, which allows/supports both free and paid access to the TV white spaces: i.e., one chamber supports free access through opportunistic or geo-location database access; and the other chamber supports paid usage through secondary spectrum trading. Consequently, four technological scenarios for the acquisition of TV white spaces emerge, namely: sensing only, joint sensing and geo-location database access, geo-location database access only, and broker based secondary spectrum trading. An analysis of these scenarios is performed based on a theoretical framework for emerging technology evaluation while considering technological, business models and regulatory dimensions. The analyses show that free and paid access to TV white space complement each other; and that despite considerable infrastructure costs, the bicameral geo-location database is positioned to create viable TV white spaces exploitation value chains; hence have the most optimal technological, business and regulatory prospects. & 2012 Elsevier Ltd. All rights reserved.

Keywords: TV white spaces Bicameral geo-location database Spectrum broker Business model Regulatory policies

1. Introduction Technological innovation for new entrants in the wireless services market such as machine-to-machine (M2M) communications, broadband Internet access, etc.; and capacity extension for current operators are the main drivers for the need of additional spectrum. TV white spaces offer a very rare opportunity to meet this demand. TV white spaces are spectrum frequency bands unused by the Digital Video Broadcasting—Terrestrial (DVB-T) systems, interleaved in both frequency and space. The exploitation of TV white spaces in different countries, however, with the exception of the USA and the UK, has not yet taken-off. It has been grounded by uncertainties regarding enabling technologies, potential business models and regulatory policies. The protection of incumbent systems is the main concern when developing cognitive radio (CR) networks operating in TV white spaces (Deb, Srinivasan, & Maheshwari, 2009; FCC, 2010; Peha, 2009). DVB-T systems and professional wireless microphones or PMSEs (Programme Making and Special Events) are considered to be incumbent users of the TV spectrum bands. PMSEs’ owners are concerned that the exploitation of TV white spaces by CRs may harm their services, which are

n

Corresponding author. Tel.: þ 351 234 377 900; fax: þ351 234 377 901. E-mail addresses: [email protected] (J.W. Mwangoka), [email protected] (P. Marques), [email protected] (J. Rodriguez).

0308-5961/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.telpol.2012.07.010

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

117

vulnerable to interference. Moreover, are DVB-T operators willing to share their bands? Most likely they are not, neither do they have incentives to improve their technologies to make sharper filters and increase the efficiency of UHF bands usage. Therefore, factors that will influence the alignment of incumbents to favor the sharing of their spectrum with CRs have to be clarified. The development of viable business models and value chains is just as important for the success of the TV white spaces usage as it is for other segments of wireless communications (Ballon & Delaere, 2009; Casey, 2009; Nolan, Mullany, Ambrose, & Doyle, 2007). What are possible business models to emerge? Moreover, an established value chain for the exploitation of TV white spaces is still missing. This is important to investors. An understanding of what factors will influence or favor certain models for making money is an important aspect for the successful usage of TV white spaces. This implies that either gradual growth or extension of existing wireless services value chains, for example towards mobile commerce or e-business value chains platforms, should be adopted. The latter approach seems to be more feasible by virtue of higher accessibility of services deployed through the Internet. Regulators are concerned about creating policies with incentives for using the white spaces while ensuring the protection of incumbent users of TV spectrum (ECC Report 159, 2011; Ghosh, Roy, & Cavalcanti, 2010; Ghosh et al., 2011; Hwang & Yoon, 2009; Peha, 2009). Questions that they have to answer include: What should be the operating parameters? How to build an accountability framework, that allows an interferer to be easily tracked down and disputes resolved either through legal or through other means? Further, how should cognitive devices be certified taking into account their ability to flexibly change operating parameters? It can be perceived that, these problems can be adequately resolved if different stakeholders in the value chain are involved or consulted. Otherwise, unilateral decisions by regulators on white space usage risks being counter-productive – either by being too restrictive rendering no technology operative, or by being too lax leading to increased harmful interference. Thus, clear identification of opportunities by involving other stakeholders can potentially balance the ecosystem and maximize economic value of the TV white spaces. As Hwang and Yoon (2008) pointed out, in addition to the research on the technology to exploit TV white spaces, it is important also to analyze the market or industry, or the proper interplay between technology, strategy and policy required to succeed in the commercialization of dynamic spectrum access (DSA) technology. Sensing technology is more appealing to academic researchers than other stakeholders (Weiss, Delaere, & Lehr, 2010); whereas regulators favor the geo-location database approach (FCC, 2010; OFCOM, 2010a; RSPG, 2011). This could be done in a commons or under spectrum trading. However, in case of scarcity, market based spectrum usage is said to be the best way to allocate spectrum resources (Cave, Doyle, & Webb, 2007; Coase, 1959; ECC Report 169, 2011; Mastroeni & Naldi, 2010), such as through Broker based secondary spectrum trading (Bae et al., 2008; Mwangoka, Marques, & Rodriguez, 2011). Therefore, a comparative analysis of the factors that might lead to one approach over another is important for optimal spectrum usage. To this end, this paper proposes a bicameral (two-chambered) geo-location database to accommodate viewpoints from diverse stakeholders, specifically the commons and secondary spectrum trading. Accordingly, four scenarios for spectrum acquisition are derived. To analyze the scenarios, this work proposes a theoretical assessment model that integrates the views from three categories of stakeholders: technology, business and regulatory. The assessment tries to answer the question whether access to TV white spaces should be free or paid, and how? The paper proceeds as follows. Section 2 introduces a bicameral geo-location database model and different approaches for accessing the TV white spaces. In addition, the section introduces a Broker based model for secondary spectrum trading of the TV white spaces. In Section 3, four scenarios are derived and a theoretical model to analyze them is presented. Sections 4–6 assess derived models from three perspectives: technological, business and regulatory, respectively. Finally, Section 7 summarizes and concludes the paper. 2. Access modes for TV white spaces 2.1. Spectrum sensing In spectrum sensing, devices try to detect the presence of protected services in each of the candidate channels. Once a vacant channel is determined, and there is assurance that interference to adjacent bands can be avoided, a device can start transmitting data in the bands. Furthermore, the device is expected to stop transmission as soon as the presence of incumbent activities is detected. The feasibility of a sensing method depends on the accuracy of the sensed data. The accuracy/reliability of sensed information can be affected by shadowing and multi-path effects prevalent in wireless channels. Consequently, even though the sensitivity of the spectrum sensors could be high, the accuracy of the spectrum availability information would be affected by the method of rendering sensed data. There are two main approaches namely (single-device) spectrum sensing and co-operative sensing. 2.1.1. Autonomous sensing Autonomous spectrum sensing is the conventional way for getting spectrum information. It is a well researched approach. However, the chaotic nature of wireless channels renders this approach (on its own) infeasible in some scenarios, especially in densely populated areas. More-over, the presence of sensing functionality in a CR increases the complexity of the device and shortens battery life.

118

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

2.1.2. Co-operative sensing By increasing the reliability of spectrum sensing information through cooperation, co-operative sensing allows devices to share their spectrum discovery resources and avoid interference with the incumbent as well as the hidden node terminal problem. The development of clever co-operative mechanisms to facilitate error-free opportunity detection is the main challenge. 2.2. Bicameral geo-location database A geo-location database is a repository for storing information on the availability of the TV white spaces. This information is recorded in form of location (or pixels), emission power levels, frequency range, time, etc. The accuracy of this information will enable efficient spectrum usage without causing harmful interference to incumbents. The geo-location database makes it possible to categorize available bands based on different criteria. This feature can be exploited to support different configurations of TV white space usage. This work envisions a bicameral geo-location database (BGDB) where a part of the TV white spaces is assigned for spectrum commons (free access) usage; and the remaining portion is assigned for secondary spectrum trading (paid access) through a Broker. Fig. 1 illustrates the bicameral geo-location database model where the spectrum commons model would operate in the bands marked for unlicensed usage, and the Broker(s) would trade the TV white spaces that are reserved for secondary trading. The spectrum commons chamber of the BGDB will serve inquiries from white space devices (WSDs) operating within the spectrum commons regime. A master-slave configuration for WSDs is envisaged: the master connects to the database and the slaves are managed by the master, i.e., they do not access the database. This configuration could reduce signaling overhead in database access. The secondary spectrum trading chamber of the BGDB will deal with inquiries coming from the Brokers. The Broker requests and receives batch data containing TV white spaces that are available for trading, and then sells them to interested parties. More details on the Broker are given in Section 2.3. Such a two chambered geo-location database will allow WiFi-like services in TV white spaces through the commons (unlicensed) usage. Additionally, it will support other services through Broker based secondary spectrum trading (i.e., licensed usage). Thus, the proposed BGDB model plays a pivotal role in supporting the development of diverse business models in the exploitation of TV white spaces. Moreover, the proposed geo-location database can accommodate two technical approaches to access the TV white spaces as follows: 1. Joint sensing and geo-location database access (Section 2.2.1); 2. Only geo-location database access (Section 2.2.2). The following subsections give the details of the two approaches. 2.2.1. Joint sensing and geo-location database access The information on DVB-T spectrum usage is stable, and that of registered PMSEs predictable. Therefore, the geo-location database can sufficiently protect such incumbents. However, the unpredictability of unregistered PMSE applications, which also

Fig. 1. A bicameral geo-location database allowing the exploitation of TV white spaces in a free or paid way through spectrum commons or secondary spectrum trading, respectively.

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

119

require protection, is the main challenge in the dependence of geo-location database. A potential solution is moving the operation of unregistered PMSEs activities to a specified band or ‘‘safe harbor’’ (ECC Report 159, 2011). However, this may take years to materialize. Therefore, a sensible near-term approach is joint sensing and geo-location database; i.e., DVB-T and planned PMSE activities are protected through the geo-location database, while unpredictable activities are protected through autonomous sensing. As shown in Section 2.1, inaccuracy and unreliability of sensed information are the major limitations of autonomous spectrum sensing. The hybrid approach allows local sensing to be performed only in a limited number of TV channels indicated by the database. In this way, the reliability of the sensed data improves. In addition, the delay due to the sensing process shortens. 2.2.2. Only geo-location database access For reliable access to TV white spaces and QoS guarantee by wireless service providers, a scenario where geo-location database access and ‘‘safe harbor’’ channels for non-registered PMSE exists is envisioned. In this scenario, sensing is assumed unnecessary for the discovery of spectrum opportunities (ECC Report 159, 2011; FCC, 2010; OFCOM, 2010a). Furthermore, it is envisioned that regulators would implement a ‘‘safe harbor’’ for exclusive PMSEs usage, i.e., a number of TV white spaces channels are reserved for PMSE usage only, in which no WSDs operation is permitted. The ‘‘safe harbor’’ bands are flexible and may vary from region to region (ECC Report 159, 2011). By confining unregistered PMSEs to ‘‘safe harbor’’ bands, WSDs can freely operate in remaining white spaces without the obligation to protect unpredictable PMSEs. This approach allows a fast re-assignment of spectrum ownership, thus facilitating QoS guarantee and increased spectrum efficiency. This scenario has the potential not only to open the market to new players, but also to create new business opportunities for the spectrum Broker entity be that in new public sector roles or in the commercial sector. For completion, the following section presents a reference Broker model. 2.3. A Broker model The spectrum Broker determines how spectrum is allocated among players, and also how much each one pays for the acquired portion (Bae et al., 2008; Berry et al., 2010; Hwang & Yoon, 2008). Formally, a Spectrum Broker (IEEE 1900.1, 2008) is an entity, device, or device capability responsible for dynamic assignment of spectrum access rights. It may lease parts of the frequency spectrum to specific parties under certain policy with or without time constraints. Further, it may be owned by the National Regulatory Authority (NRA), where applicable, or by private or independent organizations. Functionally, the Broker has the following capabilities (Mwangoka et al., 2011): Dynamic TV white space allocation; trading mechanism and price discovery; policies and occupancy repositories; and registration and validation. The following subsections give the details. 2.3.1. Dynamic TV white spaces allocation Dynamic TV white spaces allocation is designed to optimize TV white space resources allocation by implementing different optimization strategies. The TV white spaces allocation mechanism implements an algorithm that uses information from the database to determine vacant bands and power levels at which a secondary user (player) should be allowed to operate. Simultaneously, the mechanism thrives to avoid spectrum fragmentation, guarantee QoS and ensure fairness in accessing TV white spaces. 2.3.2. Trading and price discovery A player in need of (extra) spectrum buys available TV white spaces from the Broker. The trading and price discovery function facilitates TV white spaces transaction by matching the player’s requirements with available resources to achieve optimal allocation. Price of the TV white spaces being offered can be found through an auction process or by evaluating spectrum value based on the principle of opportunity cost. In this way, the Broker is able to allocate TV white spaces to the most valuable players, while maximizing the economic efficiency of the bands. The allocation mechanism must balance efficiency with complexity (Berry, Honig, & Vohra, 2010). The trading mechanism could be realized through auctioning. Alternatively, it could be achieved through the merchant approach where the Broker announces a set of reference prices for the available TV white spaces; and then adjust the prices based on time, location, bandwidth required and other factors to maximize expected revenue or to clear the market periodically. The merchant approach is simpler, and requires less information exchange overhead than the auction mechanism (Berry et al., 2010). However, a well-designed auction can achieve either higher efficiency or more revenue depending on the intended objective. 2.3.3. Policies and occupancy repositories The TV white spaces occupancy repository is the unit that contains all the information about where WSDs may transmit, which ones are active and what are their operational parameters. It also contains all the information required to estimate mutual interference between WSDs. One of the fundamental parameters in this database is spatial resolution (i.e., pixel size). Its resolution could be defined once at the introduction of the system and kept fixed, or made adaptive to different geographic regions.

120

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

Furthermore, the occupancy repository hosts the methods to manage the white space systems; and generate events (or reacts to external events) which are relevant for the management of white space systems. The repository also contains methods for populating the database; managing the database such as adding, modifying and removing TV white spaces services. Finally, the repository defines events which will trigger the updating of the database. These could be periodic or externally initiated. 2.3.4. Registration and validation Traceability is an important aspect in protecting the market from users generating harmful events. Tracing users of the Broker’s service is achieved through registration and validation. To support secure spectrum trading, a security framework is required to prevent unauthorized access to the resources. Secured transaction is an important incentive for the promotion of secondary spectrum trading. Thus, each device that accesses the TV white spaces through the Broker must be registered, otherwise it will not be allowed to operate. 2.3.5. Summary The Broker provides a platform where both the merchant and auction modes of secondary spectrum trading are supported. In both modes, the algorithm for maximizing profit and the sequences of information exchange for bidding or price adjustments are done automatically. Moreover, the Broker embodies algorithms for optimized allocation of the TV white spaces such that spectrum fragmentation is avoided, QoS is guaranteed, and fairness is ensured. To support secured transactions, the Broker provides a registration and validation mechanism to prevent unauthorized access to the resources. The security framework is an important incentive for the promotion of secondary spectrum trading. 3. Derived scenarios and theoretical assessment model This section derives four TV white spaces access scenarios and introduces a theoretical assessment framework that will be used to analyze them. 3.1. Derived scenarios From the discussion in Section 2, the following four technological scenarios for TV white spaces exploitation are derived: 1. 2. 3. 4.

S1: S2: S3: S4:

Autonomous sensing, without access to the geo-location database, in the commons; Sensing with geo-location database access, in the commons; Only geo-location database access, in the commons; Only geo-location database access, in broker based secondary spectrum trading.

The scenarios are differentiated on the basis of the technical characteristics required for accessing TV white spaces. These characteristics have direct impact on the technologies, business models and regulatory policies that will be necessary for the creation of a functional TV white space ecosystem. These scenarios generally fall within free (unlicensed) and paid (licensed) access in the commons and secondary spectrum trading based access, respectively. 3.2. Theoretical assessment model The economic value generated by the 470–862 MHz band will depend on how spectrum supply and demand are matched (Analysys Mason, 2009). Regulatory policy have direct influence on the allocation of spectrum and the conditions of spectrum use in different domains. Regulatory policy also has impact on CR technologies, hence affecting spectrum demand indirectly. Apart from technology, the main pull of demand for spectrum is the need by stakeholders to provide various wireless services to end users. Under such conditions, assessing a new technology, in our context TV white space access technology, both in terms of economic viability and future prospects, is not a straightforward undertaking. A meaningful analysis must consider the dimensions and criterion for evaluation that integrate the interests of different parties with different perspectives. Casey (2009) uses scenario planning and system dynamics methods to build scenarios for future spectrum markets considering uncertainties such as the decentralization and locality of spectrum markets and the vertical integration in the industry around the spectrum resources. Ballon and Delaere (2009) identify a number of flexible spectrum business configurations and revenue sharing models for potential cognitive pilot channel (CPC) implementations and, uses a business scorecard approach to evaluate these models. Chapin and Lehr (2007a) analyze the interactions between dynamic spectrum access (DSA) technology and the wireless services market, and recommend the path that the technology, markets, and regulations ought to follow in order to overcome potential barriers. However, these works does not particularly deal with the four technological scenarios for the access of the TV white spaces as derived in this work.

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

121

Table 1 A theoretical framework for assessment of a TV white space exploitation technology (S1/S2/S3/S4). Stakeholder

Dimension

Questions

Technology

Cost saving Interference protection Spectrum mobility Quality of service Co-existence mechanism

Does it have advantages of cost saving? Does it protected better the incumbents? Does it provide seamless spectrum mobility capabilities? Doe is it provide higher quality of service? Does it provide better co-existence with other systems?

Business models

Value proposition Competitive advantage Business partnership Innovation incentives

What value can it provide to customers? What advantage does it have over competitors ? Can it form an effective value chain for efficient spectrum usage? What incentives for innovation can it provide?

Regulatory policies

Spectrum harmonization Spectrum rights Certification

Can it enable the harmonization of spectrum usage? Can it enable the definition of clear spectrum usage rights (SURs) ? Can it enable dynamic certification of reconfigurable radio devices or equipments?

In this work, the analysis model builds upon an emerging technology evaluation methodology proposed by Benson, Sage, and Cook (1993). The work proposed that before a technology becomes useful for the society it must pass through three gateways: a market gateway, a systems management gateway and a technology gateway. In order to adopt the methodology for the telecommunications, Yuan et al. (2006) added a government policy dimension to the analysis framework. Since the purpose of TV white spaces access technology is to satisfy the demand for additional spectrum to provide wireless services; this work does not consider the market demand dimension. Rather, it considers a framework that represents and integrates three perspectives on the exploitation of TV white spaces, namely the technology perspective from technology developers, the business model perspective from the industry and the regulatory policy perspective from the NRAs. This framework is summarized in Table 1. Therefore, this work proposes that the technology, business model, and regulatory dimensions are the key perspectives to be analyzed. The analysis intends to identify which scenario for the exploitation of TV white spaces through CR technologies has the highest potential to become commercially valuable. 3.2.1. Enabling technologies The main value proposition of CR technologies is the realization of better use of radio spectrum for the society (Akyildiz, Lee, Vuran, & Mohanty, 2006; Chapin & Lehr, 2007a; Mitola, 2009), which could be achieved through any of S1 through S4 approaches. The development of new spectrum access technologies is expected to result in higher wireless service accessibility, better adaptability of devices to the local environment, improved interconnectivity among multi-terminal/ multi-frequency devices, increased scalability through collaboration, as well as improved reliability in spectrum availability among others. Considering the nature of the TV white spaces exploitation technologies, this paper analyzes technology issues in terms of five performance indicators: cost saving, interference protection, spectrum mobility, quality of service and coexistence mechanism. 3.2.2. Business models A business model is a planning tool that is essential for every company, such as a geo-location database access provider. It describes how a company creates value to customers, owners, and other stakeholders. It is also important in case the services are provided in cross-company collaboration in complex value nets, like in the TV white spaces scenarios where different entities are expected to interact in identifying, storing, and exploiting the resources. In this paper, the analysis of S1–S4 under the business models perspective will consider: value proposition, competitive advantage, and innovation incentives. 3.2.3. Regulatory policies In general, regulators represent the society to ensure maximal social utility of spectrum resources. They promote secondary spectrum usage through the policies they make, specifically the obligations of spectrum users to limit interference to other users in the market. These policies create incentives for developing and deploying innovative wireless access services, as well as enabling automated and flexible management of spectrum resources. Regulation has direct impact on innovation in two ways. First, compliance with regulations reduces the available resources for investments, leading to lower capital intensity and lower technical progress and innovation (Crafts, 2006). Second, regulation changes the incentives for investments in research and development (R&D). For example, patent protection regulations may increase the incentives to invest in R&D (Carlin & Soskice, 2006); whereas price control and product market regulations may reduce incentives to invest in R&D (Crafts, 2006). Therefore, the absence of certain regulatory policies disables the ability of investors to assess the pros-and-cons and make the trade-offs necessary for investment, for example, in TV white spaces based technologies. Under perspective, this paper analyzes S1–S4 considering: spectrum harmonization, spectrum usage rights and certification.

122

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

4. Technology perspective The technology comparisons among S1–S4 are conducted along their attributes of cost saving, interference protection, spectrum mobility, quality of service and co-existence mechanism. 4.1. Cost saving New technology development minimizes costs through more cost-efficient provisioning of old services or cost-effective ¨ provisioning of new services (Dahlman, Parkvall, Skold, & Beming, 2009). Thus, in TV white spaces, cost is incurred in spectrum acquisition (access), end user devices, infrastructure deployment, and operation. Clearly, by sharing spectrum in a commons way, S1–S3 have a big advantage over S4 in terms of spectrum access cost savings. S1 and S2 will potentially incur costs for mounting sensing capabilities on CR devices. Thus, compared to 3G devices, S1 and S2 devices could be more expensive due to the additional sensing circuitry and signal processing required. Operating costs would vary entirely based on environment: congested areas might incur higher costs compared to sparsely populated areas. Under S1, higher costs are expected in terms of investment in R&D. Further, S2 will have to trade-off signaling overhead to accuracy. S2–S4 need to invest in establishing an infrastructure for providing information on the availability of TV white spaces. This could be an Internet based geo-location database, especially since the Internet Engineering Task Force (IETF) standardizes a Protocol to Access White Space databases (PAWS); a cognitive pilot channel (CPC) standardized by the European Telecommunications Standards Institute (ETSI), or similar means. Definitely, there are similar operation costs associated with S2–S4 in terms of geo-location database management. The cost of the devices for S3 and S4 is expected to be substantially lower than those in S1 and S2. S4 requires a much more comprehensive infrastructure akin to that of an e-business platform. R&D costs could be substantially reduced by leveraging accumulated knowledge in e-business as well as adopting IP based infrastructure. From the discussion above, it is clear that S1 has a big advantage over S2–S4 in terms of infrastructure and operation cost savings. However, this conclusion can be offset by the observation made by Shapiro and Varian (1998) that information should not be evaluated according to the cost of producing it, but according to its value. Consequently, information on the TV white spaces availability should not be evaluated according to the cost of acquiring it, but according to the value it generates. In other words, it is feasible that higher costs for acquiring the TV white spaces could lead to the generation of higher value. For a better conclusion, therefore, the analysis of cost savings should be considered in conjunction with that of value proposition (Section 5.1). 4.2. Interference protection Interference refers to the effect of unwanted energy upon reception of the wanted signal, and is manifested by performance degradation, misrepresentation, or loss of information, which would not happen in the absence of that unwanted energy (Struzak, 2007). The operation of WSDs in the TV white spaces is a potential source of unwanted energy upon reception of DVB-T or PMSE systems’ signals. Regulators require that the usage of TV white spaces avoid harmful interference to incumbent users. Achieving this requires interference protection mechanisms. In essence, interference protection is the discovery of the TV white spaces, or the computation of the maximum allowed power for the interfering transmitter to avoid causing harmful interference. In S1, reducing the coverage area through emission power limitation will reduce the potentials of the hidden node problem. Further, a smaller coverage area means fewer users and overall traffic; hence, reduced relative interference leading to higher data rates. However, stringent requirement for high sensitivity in such devices both reduces their usefulness and increases their complexity (FCC, 2010; OFCOM, 2010b). S2 dissolves these challenges by lowering the high sensitivity requirement (Karimi, 2011). Assistance from the geo-location database in S2–S4 has the potential of reducing harmful interference in a feasible manner (FCC, 2010; OFCOM, 2010b; Karimi, 2011). S3 and S4 have the potential for longer range because regulators can dynamically change power emission settings in the BGDB. Controlling interference under S4 can further be enhanced through temporal, spatial and spectral parameters settings, i.e., when settling terms for temporary exclusive rights. In the long run, creating incentives for the incumbents to improve their hardware technologies such as the quality of filters (affecting, e.g., protection ratio), linearity of input amplifies, etc., is also a viable approach for reducing perceived interference and improve the efficiency of spectrum usage. This could be achieved by allowing the incumbents to deploy other value added services in the same bands, and thus incentivizing better interference protection technologies. 4.3. Spectrum mobility The availability of TV white spaces is non-harmonized across different geographic regions. Spectrum mobility can be expressed as: ‘‘If a primary user (PU) is detected in a specific portion of the spectrum in use, CR users should vacate the spectrum immediately and continue their communications in another vacant portion of the spectrum’’ (Akyildiz, Lee, &

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

123

Table 2 Spectrum mobility issues in S1–S4 under temporal, spatial and spectral domains. Scenario Temporal

Spatial

Spectral

S1

[ þ ] Faster spectrum acquisition if no (connection to) geo-location database [  ] More delay if there is congestion [  ] Energy inefficient

[ þ ] No need for network infrastructure for spectrum acquisition [  ] Not easy to manage user mobility [  ] Inability to plan user movements

[  ] Need to scan more channels for opportunity discovery [  ] Complex device [  ] Costly to aggregate spectrum

S2

[ þ ] Faster than S1 if there is congestion [ þ ] Power saving

[  ] Not easy to manage user mobility

[þ ] Scan only channels given by the geo-location database [  ] Complex device

[  ] Signaling overhead

[  ] Need geo-location database infrastructure

[  ] Delay in start-up connection to geo-location database S3

S4

[ þ ] Easy to acquire TV white spaces [ þ ] More power saving [  ] Need connection to geo-location database [ þ ] Pro-active TV white spaces availability

[  ] Hard to aggregate spectrum

[  ] Not easy to manage user mobility [  ] Need geo-location database infrastructure [ þ ] Easy to manage user mobility [ þ ] Can plan or trace user movement [  ] Costly infrastructure

[þ ] Reliable availability [þ ] Simpler devices [þ ] Possible spectrum aggregation [þ ] Reliable availability [þ ] Simpler devices [þ ] Easy Spectrum aggregation

Chowdhury, 2009, 2006). This can happen, for example, when the CR user is moving from one place to another. Spectrum mobility allows the user to discover and utilize new spectrum opportunities ‘on the fly’. Traditional mobility management mechanisms, when applied to CR networks in TV white spaces, may cause significant performance degradation because of ignoring the spectrum uncertainty issue prevalent in secondary spectrum usage (Mwangoka, Marques, & Rodriguez, 2010). Thus, technological features in S1–S4 should be evaluated considering the factors that influence connectivity in temporal, spatial and spectral domains as shown in Table 2. Therefore, the activity level of incumbent systems or the frequency of displacement of the WSD user would affect spectrum mobility in the TV white spaces. In S1, spectrum mobility depends on other factors such as hardware design and underlying sensing protocols. In S2, achieving spectrum mobility requires additional hardware and signaling overhead owing to the joint sensing and geo-location database access requirements. The overhead can lead to increased delay or service disruption during spectrum hand-over. The effectiveness of S3 in handling spectrum mobility would depend on the frequency that the WSD system (re-)consults the geo-location database. S4, through predictable and exclusive usage provides the best features to handle spectrum mobility in TV white spaces.

4.4. Quality of service The increase in portables with QoS stringent features such as video telephony (Wiegand & Sullivan, 2011) forecast the importance of QoS provisioning in wireless services operating in TV white spaces. Other examples include multimedia streaming, real-time traffic mapping and updates, etc. Consequently, whether and how (in terms of complexity and cost) an access technology can provide adequate QoS support is an important criteria for the viability and success of dynamic TV white spaces access, especially in consumer-oriented wireless systems (Kumar, Wang, Challapali, & Shin, 2011). S1–S3, under the license-exempt regime, promote efficiency through sharing, however, QoS cannot be guaranteed. Potential disruptions in connectivity in S1 will make the deployment of QoS-sensitive services infeasible. S2 and S3 are likely to perform better due to higher predictability of the availability of the TV white spaces. However, in case of congestion, the commons will experience increased mutual interference leading to degradation of QoS. S4 provides exclusivity of spectrum usage thereby more suitable for application which require QoS guarantee.

4.5. Co-existence mechanism Multiple collocated secondary networks are expected to use TV white spaces. Such systems will depict differing characteristics such as bandwidth, transmission power, system architectures, device types, etc., and must all comply with regulatory requirements to protect incumbents (Ghosh et al., 2011), and be able to function in parallel without degrading each other’s service. Achieving this requires proper co-existence mechanism.1 This has motivated several standardization 1 In literature, interference protection and co-existence mechanism are normally not clearly separable. In this work, interference protection is referring to the protection of the incumbent systems, which turns out to be the identification of the TV white spaces. Co-existence mechanism allows or enables multiple co-located white space systems to share the TV white spaces while minimizing harmful interference to each other.

124

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

efforts such as IEEE 802.22, 802.11af, 802.19 TG1, ECMA 392, ETSI RRS, etc., to develop suitable technologies for the exploitation of TV white spaces. The data transmitted in a wireless link is distorted by signals of other equipment(s) operating in the same or adjacent frequency bands. This limits the maximum data-rate achievable in a given bandwidth depending on the signal-power-tointerference-power ratio (Dahlman et al., 2009). Compared to the higher frequency ISM bands, the lower frequency TV white spaces have inherent RF propagation properties, such as, longer range and better penetration (Stavrou & Saunders, 2003). These properties are likely to benefit S4 more than the commons scenarios, i.e., S1–S3. In S1–S3, coexistence can be achieved through a specific set of etiquette rules such as ‘‘listen before talk’’ or ‘‘sense before transmit’’ or through a common control channel. Without coexistence mechanisms, many WSDs operating in the same band in close proximity would generate much interference to each other, leading to the ‘tragedy of the commons’, which can make the band unusable. S4, through centralized coordination and the assignment of temporary exclusive rights to use TV white spaces, allows the co-existence of multiple WSDs networks to operate within acceptable interference levels. Overall, S4 seems to be suited to deliver better prospects in terms of QoS, spectrum mobility, coexistence, etc. Thus, the benefits have a potential to outweigh the high cost involved in infrastructure development.

5. Business models perspective This section analyses business models of secondary spectrum acquisition from the aspects of value proposition, competitive advantage, business partnership and innovation incentives.

5.1. Value proposition Dynamic spectrum access enabled by cognitive functionalities is able to support innovation in terms of easier communications and the deployment of new applications and services more flexibly compared to the current fixed model, hence having an impact on (all) stakeholders in the telecommunications ecosystem (Nolan et al., 2007). The main value proposition of CR technologies is the realization of better use of radio spectrum for the user (Mitola, 2009), which could be achieved through any of S1 through S4 approaches. A value proposition for S1–S4 is the reason why a spectrum user would choose one method for exploiting TV white spaces and not another. The main value proposition of S1 is low-cost spectrum acquisition with limited spectrum mobility. The aspect of cost saving among S1–S4 has been discussed in Section 4.1. Users certainly want cheap spectrum, however, the stringent protection thresholds set by regulators renders sensing alone unreliable in protecting incumbents. Consequently, the low-cost spectrum acquisition in S1 would be more attractive if regulators relax sensing thresholds especially for rural areas where spectrum usage is not congested. Compared to S1, the value proposition of S2 is added reliability in sensing results due to assistance from the geolocation database to identify available bands. Nevertheless, the double sensing and signaling overhead could be costly in terms of energy saving. In commons usage, S3 offers the best value proposition through centralized geo-location database access – where no sensing is required – thus shedding off sensing related demerits without compromising reliability. The main value proposition of S4 is exclusive TV white space usage where wireless access services with guaranteed QoS can be deployed. Another value proposition of S4 is the support for spectrum mobility. According to Shapiro and Varian (1998), ‘‘the technology infrastructure makes information more accessible and hence more valuable’’. The Broker (S4), therefore, is a value generator for efficient spectrum usage and innovative business models. It is not merely a tool for enabling secondary spectrum trading, but rather, an introduction of the use of ICT (Information & Communication Technologies) in the regulation, management, and allocation of spectrum. The same applies for the geo-location database in S2 and S3. The ongoing standardization work in IETF PAWS can be regarded as a precursor for the value enhancement of S2–S4 models.

5.2. Competitive advantage Competitive advantage outlines the points that make a spectrum access technology more attractive to potential users compared to other options. The question posed, is why opt for TV white spaces, and under which scenario? Different options offer different competitive advantages as perceived by different users. Secondary spectrum users or buyers could be new entrants to the wireless service market or established service providers. When they need (extra) spectrum to support a specific service with multi-dimensional requirements (including coverage, bandwidth, QoS, access duration, price, etc.), there are several options to obtain the needed resource. These include optimization of own spectrum, purchase of new spectrum with exclusive rights and, off course, the use of TV white spaces (Mwangoka et al., 2011). In order to effectively evaluate competitive advantage; alternatives competing against the usage of TV white spaces are given first in the following subsections.

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

125

5.2.1. Optimization of current own spectrum Service providers may improve the efficiency of the spectrum they own through introducing better frequency planning, create better coexistence strategies or re-assignment. This is made possible through initiatives that allow technology neutral usage of spectrum for example (RSPG, 2005; DG INFSO, 2006). This option is not always possible since current wireless networks are usually spectral efficient. Further, this option is not feasible to new entrants since they are considered not to have spectrum resources. Therefore, TV white spaces offer extra spectrum and low entrance barrier for both service providers and new entrants, respectively. 5.2.2. Purchase of new spectrum with exclusive rights The service provider that envisions a nationwide market dimension to leverage economies of scale and international harmonization is more likely to buy new spectrum with exclusive rights, e.g., the digital dividend band (790–862 MHz) that has been released after the analogue TV switch off. This may lead to the ‘‘Winner’s Curse’’ phenomenon in case the bid winner with incomplete information overpays to obtain spectrum and encounters poor market response to the services offered. Moreover, this approach may be prohibitively costly to new entrants who are normally constrained in terms of capital. In other words, TV white spaces have lower chances for the ‘‘Winner’s Curse’’, and are much less capital-intensive. 5.2.3. The use of TV white spaces TV white spaces offer several options both for existing service providers who want to increase their capacities or coverages and for new entrants who want to venture into innovative wireless services markets. TV white spaces can be obtained in terms of temporarily exclusive rights purchased from a secondary market or through unlicensed commons usage. S1–S3, is directed towards users who are satisfied with best effort services such as WiFi-like technologies, which operate under the commons. S4, is directed towards users who demand QoS guarantee. Therefore, operators interested in deploying services offering QoS are considered to be the main beneficiaries of S4. Such systems are to benefit from extension to the acquired bands through increased radio coverage and system capacity, whereas exclusivity of spectrum use will ensure that the constraints imposed by the SLA (Service Level Agreement) between the network operators/service providers and the users are consistently achieved. 5.3. Business partnership Successful exploitation of TV white spaces will depend on the establishment and maturity of value chains (Bentz, Delgado, Jain, Kakodkar, & Salomon, 2011), such as in Fig. 2. Top in the value chain are the regulators whose intent it to maximize the social value of the spectrum. Next are the intermediary spectrum managers, such as the geo-location database managers, Brokers, Band Managers, etc. They facilitate dynamic spectrum access in a free or paid manner depending on their revenue models. Following are the hardware and software makers which develop the necessary chips and devices, including spectrum sensors, reconfigurable mobile devices, white space access points or base stations, etc. Next, network operators to provide the necessary infrastructure, and then the service providers to provide wireless access services, spectrum mobility management, etc. Finally, the end users which could be extended WiFi users, M2M communications, smart grids, public safety systems, etc. As it can be seen in Fig. 2, the intermediary spectrum managers, the geo-location database managers and Brokers in this case plays a pivotal role in the TV white spaces exploitation value chain. Therefore, S2–S4 may leverage existing Internet based value chains to offer TV white space provision services. The exclusive TV white spaces access feature in S4 implies that spectrum users such as service providers will have to sign some sort of contracts with the Brokers. Obviously, such service providers would have opted out free spectrum access through S1–S3. Moreover, S2 and S3 may provide TV white space availability information for free, but generate revenue through advertisements. The sensing functionalities in S1 and S2 need to gain the support of Hardware Makers, whereas geo-location database access support in S2–S4 have to be considered by Software Developers. Moreover, spectrum sensing in S1 and S2 can be employed as a value adding service for a white space provision service (Weiss et al., 2010), such as in S3 and S4. In summary, the exploitation of TV white spaces is in its infancy, and there are several potential configurations of viable business models, suggesting that the respective value chain would occur along multiple dimensions. However, the current regulatory efforts in adopting the geo-location database for example, the (OFCOM, 2010a; RSPG, 2011), as well as related standardization efforts such as those by the IETF PAWS, indicate that S3 and S4 have a high potential of creating a viable value chain for efficient spectrum usage, for example, by extending conventional mobile e-commerce value chains into the TV white spaces. Intermediar Regulators y Spectrum Managers

Hardware Makers

Software Developers

Network Operators

Service Providers

Fig. 2. A potential TV white spaces exploitation value chain.

End Users

126

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

5.4. Innovation incentives TV white spaces can support different types of use cases such as LTE, WiFi, Public Safety, M2M communications, etc. Also, these systems support different topologies including cellular for LTE, ad hoc and infrastructure for WiFi and Mesh for Public Safety systems. The details are as follows: 5.4.1. Innovation in the commons bands (S1–S3) WiFi is one of the high potential use cases for TV white spaces. WiFi is a low cost mature technology which can be benefited by migrating from 2.4 GHz to the lower UHF range. Range extension would be the key incentive for deploying WiFi in the TV white spaces. If one considers the simplified scenario where the free space loss formula can be used, usage of the 2.4 GHz band suggests a maximum transmission range of 250 m. However, if the service is migrated to the 500 MHz TV spectrum then the maximum transmission range becomes 1.2 km, which suggests a significant range benefit at the maximum sustainable connection rate of 54 Mbps. IEEE 802.11af is a standard that extends WiFi to the TV white spaces. WiFi can operate in the commons bands, following the ISM bands usage model. However, in S4, an operator interested in providing premium WiFi access services with cellular like security and high QoS guarantee, may acquire bands from the market for the targeted high-value customer base. 5.4.2. Innovation in secondary spectrum trading (S4) The secondary spectrum market proposed is a suitable regime to guarantee QoS for cellular systems whenever they extend their services to the TV white spaces. Here an example of LTE system is given. By using TV white spaces for LTE, it is feasible to reduce the number of base stations, while offering the same or even higher throughputs (Mwangoka et al., 2011). This is very important to both the network operator and the users. The network operator decreases the CAPEX (Capital Expenditures) and OPEX (Operational Expenditures). The end user experiences better services. However, these merits should be realized while the constraints imposed by the Service Level Agreement (SLA) between the network operators or service providers and the user – defining the minimum QoS – are not compromised. Thus, by ensuring the exclusivity and low interference levels, S4 offers a QoS guarantee incentive for LTE systems. Furthermore, the extension of LTE over TVWS can be facilitated by the use of the carrier aggregation feature where multiple LTE ‘‘component carriers’’ are aggregated on the physical layer to provide the necessary bandwidth (Parkvall et al., 2008). This feature can be achieved through a Broker based model through trading. 5.4.3. Innovation in e-Regulatory management (S2–S4) The exploitation of TV white spaces through CR technologies opens up the need for automated regulatory management. For example, in S2–S4, the operation of the geo-location database requires the use of information communication technologies (ICT) to enable efficient management of spectrum allocation. Furthermore, in S4, secondary spectrum trading, through the Broker for example, give incentives for the adaptation of e-business models to enable efficient and secure exchange of spectrum rights. In addition, these dynamics pose a challenge in the certification of reconfigurable radio systems. In that regard, a dynamic mechanism is needed to ensure the conformity of devices to regulatory requirements. Specifically, a trust network across various stakeholders to support automated certification should be established. More discussion of dynamic conformity is given in Section 6.3. 6. The regulatory policies perspective In this section, the viability of regulatory policies in S1–S4 is assessed in relation to spectrum harmonization, spectrum usage rights and certification of reconfigurable radio devices operating in TV white spaces. 6.1. Spectrum harmonization The excellent propagation and coverage characteristics of TV white spaces have attracted major stakeholders who will definitely be interested in harmonized spectrum usage across different countries for two reasons. First, harmonization would enable standardization of (spectrum access) technologies which in turn would lead to the achievement of economies of scale. Second, harmonization of spectrum usage would facilitate the movement of devices from one country to another. In that regard, the key to achieve regulatory harmonization in spectrum usage is technical standardization. In S1–S4, technical standards for the radio access technology to use TV white spaces as well as the communication protocols between various stakeholders will be eminent. Currently, there are a lot of efforts to standardize sensing based (i.e., S1 and S2) radio access technologies to operate in TV white spaces under the auspices of IEEE 802.22, IEEE 802.11af, Weightless SIG, ETSI RRS, etc. In the case of S3 and S4, the IETF PAWS (Protocol to Access White Space database) is specifying messaging interfaces between devices and white space databases under different conditions. Furthermore, different regions would have different methods to define the availability of TV white spaces as well as related maximum emission power limits. In this case, devices operating under the S1 regime, would have difficulties in

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

127

moving from one region to another, as pointed out in Section 4.3. For S2, information on maximum emission power limits can be obtained from the database—thus has the ability to operate in different regions. S3 and S4 are better suited to exploit flexibility in defining methods and parameters alike. Moreover, S2–S4 allows regulators to change their policies dynamically to maximize spectrum usage efficiency (see also Section 5.4.3). Lastly, when secondary spectrum trading is involved (as in S4), the definition of the interfaces between the Broker and the white space database should be flexible to allow country specific requirements regarding trading and spectrum rights. Cross-border issues could be easily addressed under S3 and S4 through the white space geo-location database.

6.2. Spectrum usage rights Appropriate definition of spectrum usage rights is presumed when spectrum is shared through a market-based mechanism. Different regulators use different aspects to characterize spectrum usage rights (Marcus et al., 2005). Functionally, a spectrum usage right (SUR) is a license which limits the interference that an owner (licensee) is allowed to cause in terms of signal strength as experienced by other owners (or licensees) (Bae et al., 2008; Cave & Webb, 2011). Signal strength can be measured in terms of power flux density which expresses the power at a certain point in terms of watts/m2/MHz . A clearly defined SUR gives substance to an intangible spectrum good. In principle, S1–S3 operate under the commons model of TV white spaces usage. The benefits include ‘‘low entry barriers, certainty of obtaining access, lower demand for licensed spectrum, innovation (anti-monopoly), public infrastructure, freedom of speech/cultural diversity, etc.’’ (Pagorel, 2007). Thus, in the context of S1–S3, defining spectrum policies and etiquette rules to avoid the ‘‘tragedy of the commons’’ as well as guarantee QoS in the commons usage is an alternative to defining SUR. Currently, there is an ongoing standardization work for coexistence mechanisms in the TV white spaces done by the IEEE 802.19 Task Group 1 (IEEE 802.19.1). The aim is to develop coexistence mechanisms that allow two or more devices or networks to efficiently exist together at the same time, in the same place, and operate in the same frequency band, without causing each other harmful interference. Particularly in S1 and S2, ‘spectrum usage rights’ could be achieved in terms of incentives to encourage cooperative behavior, for example a device relaying other devices’ traffic in exchange for using their spectrum, or giving information on TV white spaces availability. S4 provides features that support the basic conditions necessary to establish a SURs based business model for secondary spectrum trading. Such features include: a platform for secondary markets, clearly defined rules and enforcement mechanism to enable risk free SURs transactions among others (Bae et al., 2008; Hwang & Yoon, 2009; Xavier & Ypsilanti, 2006). Thus, S4 makes the introduction of SURs in secondary spectrum trading feasible.2

6.3. Certification A flexible PHY architecture enables reconfigurable wireless terminals to operate in heterogeneous networking environments. In order to function seamlessly across various markets, radio equipments have to be certified. This flexibility brings new challenges in enforcing device certification and fostering favorable conditions for the introduction of new innovative applications (Chapin & Lehr, 2007b). In additional to device flexibility, differing policies are also an important challenge. For example, some regulators would not permit any changes to the device once licensed (Suzuki et al., 2003). Therefore, a manufacturer or users must obtain a new license once the license terms have been changed. In the EU context (R&TTE DIRECTIVE, 1999): certification regulation requires self-declaration of conformity of a given equipment by the manufacturer or the importer into the common market. Furthermore, the EU is considering updating the R&TTE (Radio equipment and Telecommunications Terminal Equipment) Directive to account for dynamic declaration of conformity for reconfigurable radio systems (Mueck, Haustein, & Bender, 2011). In the USA, however, the FCC controls both spectrum allocation and device certification for reconfigurable devices before entering the internal market. Thus, differing policies pose a challenge in achieving seamless connectivity of reconfigurable radio systems while conforming to regulatory requirements. Accordingly, a dynamic mechanism for ensuring the conformity of devices to regulatory requirements, with certainty, has to be developed. This can be achieved by establishing a trust mechanism across various stakeholders to ensure seamless and dynamic conformity. In this light, S1 as a device centric technology for spectrum access is more challenging to achieve dynamic certification. On the other hand, S2–S4, which are network-centric technologies for spectrum access have higher viability to operate in a dynamic conformity environment. 2 The RSPG (2011) has introduced the concept of ‘‘Licensed Shared Access’’ (LSA), defined as: an individual licensed regime of a limited number of licensees in a frequency band, already allocated to one or more incumbent users, for which the additional users are allowed to use the spectrum (or part of the spectrum) in accordance with sharing rules included in the rights of use of spectrum granted to the licensees, thereby allowing all the licensees to provide a certain level of QoS. S4 has the capability to enable the realization of LSA through the provision of SURs.

128

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

7. Summary and conclusion This work has attempted to analyze the technological, business and regulatory factors that will enable the exploitation of TV white spaces. A bicameral (two chambered) geo-location database, which supports both free and paid usage of TV white spaces under the commons and secondary spectrum trading regimes, respectively, was proposed. Four scenarios, namely; autonomous sensing, joint sensing and geo-location database, only geo-location database and Broker based secondary spectrum trading; were identified and analyzed by using an emerging technology evaluation methodology that integrates the perspectives of technology, business and regulatory stakeholders. Under the technology perceptive, the four scenarios were analyzed in terms of five performance indicators: cost saving, interference protection, spectrum mobility, quality of service and coexistence mechanism. Overall, the fourth scenario, i.e., Broker based secondary spectrum trading, seemed to have better prospects in terms of interference protection, spectrum mobility, quality of service and co-existence mechanism. These benefits have the potential to outweigh the significant cost incurred in terms of infrastructure development. The analysis of the four scenarios under the business models perspective considered value proposition, competitive advantage, and innovation incentives. The four scenarios depict differing potentials for building business models. However, the potential for developing value chains was seen as a key determinant for the success of a TV white spaces exploitation technology. Supported by outlooks from the regulatory and business model realms, scenarios based on the geo-location database appeared to have a better value proposition, competitive advantage, and innovation incentives. Under the regulatory policies perspective, the four scenarios were analyzed considering spectrum harmonization; spectrum usage rights (SURs) and certification. Automated regulatory management is regarded as a key enabler for the exploitation of TV white spaces through cognitive technologies, because of its capability to facilitate spectrum harmonization, definition of SURs and dynamic certification of reconfigurable radio devices/equipments. Thus, compared to autonomous sensing technology, geo-location based technologies seem to have the most optimal prospects. Whereas regulators are cautious to allow the usage of TV white spaces in general, this analysis shows that exploitation based on the proposed bicameral geo-location database that separates bands for free and paid usage should be adopted. Concurrently, the invocation of ‘‘safe bands’’ for the operation of unregistered PMSEs is inevitable to allow for better coordination of TV white spaces usage for the benefit of both incumbents and secondary spectrum users. This is further augmented by the potentials for developing viable value chains as shown in this study. Although, under scarcity, market based spectrum usage is preferred to commons usage, this analysis shows that they complement each other in the TV white spaces. Both can be used to deliver wireless services with different QoS provisioning strategies: best-effort services fit under the commons, while services requiring QoS guarantee fit under the secondary spectrum trading regime. Consequently, free and paid usage of TV white spaces will lead to the development of different business models, access technologies and regulatory policies. Finally, adoption of appropriate technical, business and regulatory approaches for the exploitation of TV white spaces will incentivize or motivate incumbent users of TV spectrum to allow other players in the bands. Examples of incentives to incumbents could include technologies that guarantee non-interfering operation to incumbents’ services; business models that connect incumbents to the TV white spaces exploitation value chain; and regulatory policies that allow flexible spectrum usage such that even incumbents could use TV white space bands for other services as well. How incentive schemes for the exploitation of TV white spaces should work is an important topic for further study.

Acknowledgments The authors would like to thank the anonymous reviewers and special issue editors for providing constructive comments and suggestions that helped to improve the quality of the paper. The research leading to these results has received funding from the European Community’s Seventh Framework Programme [FP7/2007-2013] under grant agreement no. 248560 [COGEU]. References Akyildiz, I. F., Lee, W.-Y., & Chowdhury, K. R. (2009). CRAHNS: Cognitive radio ad hoc networks. Ad Hoc Networks, 7, 810–836. Akyildiz, I. F., Lee, W.-Y., Vuran, M. C., & Mohanty, S. (2006). NeXt generation/dynamic spectrum access/cognitive radio wireless networks: a survey. Computer Networks, 50, 2127–2159. Analysys Mason, H., & DotEcon, H. (2009). Exploiting the digital dividend’ – a European approach. Report to the European Commission. Bae, J., Beigman, E., Berry, R., Honig, M., Shen, H., Vohra, R., & Zhou, H. (2008). Spectrum markets for wireless services. In Third IEEE symposium on new frontiers in dynamic spectrum access networks (DySPAN ) 2008 (pp. 1–10). Ballon, P., & Delaere, S. (2009). Flexible spectrum and future business models for the mobile industry. Telematics and Informatics, 26, 249–258. Benson, B., Sage, A., & Cook, G. (1993). Emerging technology-evaluation methodology: With application to micro-electromechanical systems. IEEE Transactions on Engineering Management, 40, 114–123. Bentz, N., Delgado, V., Jain, D., Kakodkar, V.,& Salomon, E. (2011). White spaces: An overview of white spaces over the next 3–5 years, evaluating potential applications from a technology, regulatory and business perspective. Prepared for the Wharton Spring 2011 Enabling Technologies Seminar. Berry, R., Honig, M., & Vohra, R. (2010). Spectrum markets: motivation, challenges, and implications. IEEE Communications Magazine, 48, 146–155. Carlin, W., & Soskice, D. (2006). Macroeconomics: imperfections institutions and policies. Oxford University Press. Casey, T. (2009). Analysis of radio spectrum market evolution possibilities. Communications & Strategies, 3 Quarter, 109–130.

J.W. Mwangoka et al. / Telecommunications Policy 37 (2013) 116–129

129

Cave, M., Doyle, C., & Webb, W. (2007). Essentials of Modern Spectrum Management. Cambridge University Press. Cave, M., & Webb, W. (2011). The unfinished story of usage rights for spectrum. In DySPAN ’11: IEEE symposium on new frontiers in dynamic spectrum access networks (pp. 41–46). Chapin, J., & Lehr, W. (2007a). COGNITIVE RADIOS FOR DYNAMIC SPECTRUM ACCESS – the path to market success for dynamic spectrum access technology. IEEE Communications Magazine, 45, 96–103. Chapin, J., & Lehr, W. (2007b). Time-limited leases in radio systems. IEEE Communications Magazine, 45, 76–82. Coase, R. H. (1959). The federal communications commission. Journal of Law and Economics, 2, 1–40. Crafts, N. (2006). Regulation and productivity performance. Oxford Review of Economic Policy, 22, 186–202. ¨ Dahlman, E., Parkvall, S., Skold, J., & Beming, P. (2009). 3G evolution: HSPA and LTE for mobile broadband (2nd ed.). Elservier Academic Press. Deb, S., Srinivasan, V., & Maheshwari, R. (2009). Dynamic spectrum access in DTV whitespaces: Design rules, architecture and algorithms. In MobiCom ’09: Proceedings of the 15th annual international conference on Mobile computing and networking (pp. 1–12). DG INFSO (2006). Mandate to CEPT to develop least restrictive technical conditions for frequency bands addressed in the context of WAPECS. ECC REPORT 159 (2011). Technical and operational requirements for the possible operation of cognitive radio systems in the white spaces of the frequency band 470–790 MHz. CEPT. ECC REPORT 169 (2011). Description of practices relative to trading of spectrum rights of use. CEPT. FCC (2010). Second memorandum opinion and order. Ghosh, C., Roy, S., & Cavalcanti, D. (2011). Coexistence challenges for heterogeneous cognitive wireless networks in TV white spaces. IEEE Wireless Communications, 18, 22–31. Hwang, J., & Yoon, H. (2008). Dynamic spectrum management policy for cognitive radio: An analysis of implementation feasibility issues. In DySPAN ’08: 3rd IEEE symposium on new frontiers in dynamic spectrum access networks (pp. 1–9). Hwang, J., & Yoon, H. (2009). A mixed spectrum management framework for the future wireless service based on techno-economic analysis: The Korean spectrum policy study. Telecommunications Policy, 33, 407–421. IEEE 1900.1 (2008). IEEE standard definitions and concepts for dynamic spectrum access: Terminology relating to emerging wireless networks, system functionality, and spectrum management. IEEE Std 1900.1–2008 (pp. c1–48). Karimi, H.R. (2011). Geolocation databases for white space devices in the UHF TV bands: Specification of maximum permitted emission levels. In DySPAN ’11: IEEE international symposium on new frontiers in dynamic spectrum access networks (pp. 443–454). Kumar, A., Wang, J., Challapali, K., & Shin, K. (2011). Analysis of QoS provisioning in cognitive radio networks: A case study. Wireless Personal Communications, 57, 53–71. Marcus, J.S., Nett, L., Scanlan, M., Stumpf, U., Cave, M., & Pogorel, G.(2005). Toward more flexible spectrum regulation. Wik-Consult Study for the Federal Network Agency. Mastroeni, L., & Naldi, M. (2010). Spectrum trading: An abstracted bibliography. CoRR, abs/1003.2554. Mitola, J. (2009). Cognitive radio architecture evolution. Proceedings of the IEEE, 97, 626–641. Mueck, M., Haustein, T., & Bender, P.(2011). Digital and dynamic certification in the framework of the novel revised R&TTE directive in Europe. In CROWNCOM ’11: 6th international conference on cognitive radio oriented wireless networks and communications (pp. 1–5). Mwangoka, J., Marques, P., & Rodriguez, J. (2011). Exploiting TV white spaces in Europe: The COGEU approach. In DySPAN ’11: IEEE symposium on new frontiers in dynamic spectrum access networks (pp. 608–612). Mwangoka, J.W., Marques, P., & Rodriguez, J. (2010). Cognitive mobility management in heterogeneous networks. In MobiWAC’10: Proceedings of the8th ACM International symposium on mobility management and wireless access (pp. 37–44). Nolan, K., Mullany, F., Ambrose, E., & Doyle, L. (2007). Value creation and migration in adaptive cognitive and radio systems. In H. Arslan (Ed.), Cognitive radio, software defined radio, and adaptive wireless systems (pp. 145–159). Springer (chapter 5). OFCOM (2010a). Implementing geo-location. OFCOM (2010b). SRSP: The revised framework for spectrum pricing - proposals following a review of our policy and practice of settling spectrum fees. OFCOM. Parkvall, S., Dahlman, E., Furuskar, A., Jading, Y., Olsson, M., Wanstedt, S., et al. (2008). LTE-advanced – evolving LTE towards IMT-advanced. In VTC’08Fall: IEEE 68th vehicular technology conference (pp. 1–5). Peha, J. (2009). Sharing spectrum through spectrum policy reform and cognitive radio. Proceedings of the IEEE, 97, 708–719. Pogorel, G. (2007). Nine regimes of radio spectrum management: A 4-step decision guide. Communications & Strategies, 1st Quarter, 169–193. RSPG (2005). Opinion on wireless access policy for electronic communications services (WAPECS) (a more flexible spectrum management approach). RSPG 05-102 FINAL. RSPG (2011). Report on CUS(collective use of spectrum) and other sharing approaches. RSPG 11-392 DRAFT. R&TTE DIRECTIVE (1999). On radio equipment and telecommunications terminal equipment and the mutual recognition of conformity. Shapiro, C., & Varian, H. R. (1998). Information rules: A strategic guide to the network economy. Boston, MA, USA: Harvard Business School Press. Stavrou, S., & Saunders, S. (2003). Factors influencing outdoor to indoor radio wave propagation. In ICAP’03: 12th international conference on antennas and propagation (Vol. 2, pp. 581–585). Struzak, R. (2007). Interference & co-existence. ICTP-ITU School on Wireless Networking for Development, The Abdus Salam International Centre for Theoretical Physics ICTP. Suzuki, Y., Oda, K., Hidaka, R., Harada, H., Hamai, T., & Yokoi, T. (2003). Technical regulation conformity evaluation system for software defined radio. IEICE Transactions on Communications, E86-B, 3392–3400. (Special Issue on Software Defined Radio Technology and its Applications). Weiss, M.B.H., Delaere, S., & Lehr, W.H. (2010). Sensing as a service: An exploration into practical implementations of DSA. In DySPAN’10: IEEE international symposium on new frontiers in dynamic spectrum access networks (pp. 1–8). Wiegand, T., & Sullivan, G. (2011). The picturephone is here. Really. IEEE Spectrum, 48, 50–54. Xavier, P., & Ypsilanti, D. (2006). Policy issues in spectrum trading. INFO, 8, 34–61. Yuan, Y., Zheng, W., Wang, Y., Xu, Z., Yang, Q., & Gao, Y. (2006). Xiaolingtong versus 3G in China: Which will be the winner?. Telecommunications Policy, 30, 297–313.