|
The radio spectrum is one of the key attributes for the success of the future of radiocommunications. The radio spectrum is a valuable commodity, and a unique natural resource shared by various types of service. Unlike other natural resources, it can be repeatedly re-used, provided certain technical conditions are met. In practice though, it is finite, can only accommodate a limited number of simultaneous users, and requires careful planning and management to maximize its value for all services. This is especially true since the demand for communication spectrum worldwide is increasing rapidly. Within this context, traditional ways to assess the merits of new technical solutions and to allocate as well as to tax use of frequencies appear to be less and less well adapted. Consequently, efficient use of the spectrum (i.e. ensuring that new spectrum is only assigned when really needed) is, together with the introduction of new technologies, the major challenge. To get more capacity, one could try to use the existing spectrum more efficiently, but with the penalty of increased infrastructure costs. A better approach is to acquire more spectrum so that each user can transmit in wider bandwidths, but this may be to the detriment of other users. Obtaining more frequency bands for mobile communications requires complex negotiations at the ITU's World Radiocommunications Conference (WRC). The process is usually protracted, and this necessitates forward planning by spectrum regulators. This paper describes some of the issues and paradigms associated with a "Dynamic Spectrum Management" (DSM) framework, from the engineering/technology, economic and radio policy aspects. It focuses on new broadband technologies that represent a challenge for the current rigid spectrum regulation. New spectrum allocation practices
The current spectrum regulatory framework is based on the allocation of the spectrum to Radio Services (e.g. Fixed Service, Mobile Service). This is reflected in the Radio Regulations (RR) [1] published by the International Telecommunication Union (ITU), which contains definitions of these Services, and a table defining their allocations for each of the ITU Regions. An example of allocation is depicted in Table 1. 
Even if RR includes mechanisms allowing some flexibility, it becomes more and more difficult for new technologies to be brought to market in time with sufficient spectrum. Therefore several ways to increase the flexibility of this framework are envisaged to better satisfy market requirements [2-3], e.g. the definition of flexible bands. Nevertheless, this evolution is made difficult by the need to protect the interests of incumbent spectrum users, and the scarcity of resources in the most promising bands. Service-neutral versus service-specific approaches
The general trend towards more flexible spectrum management is driven by the ever-growing market pressure for more commercial spectrum, and the continuous appearance of new technologies offering a wide range of applications [4]. However, a reasoned approach is needed during the transition period to meet market requirements in those bands where and when product solutions become available, and to protect solutions in those bands where exclusive rights are needed to guarantee quality of service. The greater the flexibility in spectrum licensing conditions, the higher the risk of interference. The right balance between technical efficiency, economic efficiency, and flexibility/ neutrality in spectrum usage enabling innovative product solutions has to be found. Several issues need to be addressed in a coordinated way in order to determine the conditions and requirements for specific spectrum allocations. One approach under consideration is based on adopting either: - "Technology-neutral" spectrum allocations to allow new technologies to be deployed in strictly conditioned spectrum, e.g. Time Domain Duplex (TDD) in Frequency Domain Duplex (FDD) bands.
- "Service-neutral" spectrum allocations in cases where legacy restrictions (service definitions - Fixed vs. Mobile vs. Satellite vs. Broadcast services) could hamper full use of market-critical product features (e.g. "full mobility" for WiMAX products in fixed wireless access bands). To fully benefit from these technologies would require either additional radio service allocations in the targeted bands - a time-consuming process - or a neutral approach in the usage modes of the band, e.g. mobile usage in a fixed allocation, provided that it is validated by compatibility studies based on reasonable market and technical assumptions.
- Protection for "harmonized" spectrum allocations based on exclusivity rights or strict sharing conditions, in particular those intended for nationwide/international coverage (satellite/mobile applications).
Note that one recent example of a service-neutral approach to spectrum regulation is the proposed "Wireless Access Policy for Electronic Communications Services" (WAPECS) initiative in Europe, which aims to identify a basket of frequency bands that can be allocated on a technology- and service-neutral basis, while satisfying all technical requirements against interference. This approach is expected to offer an effective and efficient use of spectrum while avoiding any distortion of competition [5]. A similar approach, identified by the UK Regulator OFCOM, considers these issues in terms of applying either technology-based, usage-based or spectrum mask licensing restrictions: see Table 2 for a summary of issues presented in the OFCOM report [6]. 
Flexibility versus harmonization
Balancing between flexibility and harmonization is not a simple task, and the benefits of either approach should be carefully considered on a case-by-case basis [7]. Flexibility facilitates the timely introduction of new applications, but presents a risk of market fragmentation, and complicates interference scenarios and co-existence issues. Conversely, harmonization favors CAPEX savings for industry, but could also freeze the use of a band, if spectrum needs evaluations for a specific usage are too optimistic. Generally speaking, new applications generally require more flexibility, while incumbent applications can tolerate a more conservative approach. The bands identified for WAPECS, which are candidate flexible bands, will be analyzed carefully to highlight the benefits and disadvantages of flexibility in their usage. Licensed versus unlicensed bands
Three cases are identified: - Applications that require a license to operate (broadcast, mobile systems, satellites, etc.). Spectrum licensing was historically the normal way to use spectrum.
- Applications for which a license is not required: industrial equipment, medical equipment, Wireless Local Area Networks (WLANs), alarms, cordless phones, etc. This case is formally known as "license-exempt", usually referred to as "unlicensed", and it is also covered by "Short Range Device" (SRD) regulations.
- Applications operating under a "light licensing" regime on a first-come, first-served basis, with the user notifying the regulator of the position and characteristics of the radio station, and new users obliged to reach an agreement with existing users in case interference criteria are exceeded. This intermediate situation is suggested for WiMAX in the 5.8 GHz band, or microwave links in bands above 55 GHz.
Several issues will need careful tracking: - Transfer of licensed bands to other regimes: the increasing popularity of unlicensed systems, particularly WiFi, may lead to calls for new and/or expanded allocations of license-exempt spectrum. Attempts to transfer existing licensed bands to unlicensed operation will have a direct impact on existing systems designed for these bands.
- Changes to operating rules for license-exempt equipment: equipment designed to operate in the license-exempt bands are subject to standards for Short Range Devices (SRD). Issues to watch are changes to tolerated emission levels, and rules governing interference avoidance procedures such as dynamic frequency selection, etc.
Paired vs. unpaired allocations
Radio systems offering bidirectional telecommunications services (uplink and downlink) can be designed to use either Time Division Duplex (TDD), where the separation between the two transmission directions is expressed in time, or Frequency Division Duplex (FDD), where transmit and receive signals are sent simultaneously, and their separation ensured by the use of different frequencies. The duplex method is often a design assumption for a given system, with examples such as GSM and satellite systems using FDD and hence assigned paired frequency bands, while WiFi uses TDD operating in a single frequency block (known as an unpaired band). There are, however, applications where both duplex modes are supported: WiMAX supports both TDD and FDD modes for fixed operation; and UMTS includes both an FDD (WCDMA) and TDD (TD-SCDMA, officially known as Low Chip Rate LCR-TDD, and IP Wireless's High Chip Rate, HCR-TDD) mode. The use of FDD and/or TDD becomes an issue for spectrum management, since FDD requires a paired allocation of two frequency bands (for up- and downlinks), while TDD systems can operate using both paired and unpaired allocations. Potential issues therefore occur whenever new spectrum allocations or assignments are to be made, hence whether or not a paired band must be found, or if an existing allocation is to be modified to permit TDD operations. In both cases, the definition of rules concerning co-existence of FDD and TDD systems in the same or adjacent frequency bands is essential. In all cases, the technical solution is to allocate sufficient guard band spacing between systems, and/or ensure sufficient distance between base stations, and/or add sufficient filters to transmitters and receivers. As the problem occurs when a new TDD system is installed near to an existing FDD system, the burden is carried by the TDD system. Coordination measures between operators, e.g. synchronization of TDD networks, will restrict interference scenarios between TDD systems to those occurring between FDD systems. Impact of future mitigation tools
All means used to minimize interference between radio systems are designated as "mitigation techniques". Up to now these were generally basic operational restrictions (e.g. power restrictions, geographical separation), or basic system features like Automatic Transmit Power Control (ATPC). Recently the use of more sophisticated techniques emerged for unlicensed applications: These techniques can be classified as: - those that forbid simultaneous use of the same frequency where this results in an interference case, e.g. Listen Before Talk (LBT);
- those that minimize the occurrence of such interferences, e.g. Low Duty Cycle (LDC).
Similar techniques are envisaged in other areas: standards are under development for access systems to ensure co-existence with broadcasting applications in the UHF band. Access to new bands
Future spectrum requirements are highly dependent on the demand for cellular services and the traffic they generate. The UMTS Forum and the European Commission (EC) have commissioned studies to investigate possible future demand for these services and the resulting traffic1. These studies concluded that the total spectrum needs of mobile systems by 2020 will be between 1.6 GHz and 2.6 GHz. This includes the current total allocation for all mobile systems (2G, 3G, DECT, extension bands, etc.) of approximately 585 MHz. Using the lower estimate of future systems of 1.6 GHz, the above studies reveal that mobility needs could require an additional amount of spectrum of around 1GHz in the future. Based on these inputs, ITU-R WP8F also conducted a study, and its recent report (May 2006) on spectrum estimates reveals that lower market assumptions give a total spectrum requirement of 1280 MHz, while higher market assumptions result in 1720 MHz, i.e. quite close to UMTS Forum estimates. Agenda Item 1.4 of WRC2007 clearly "considers frequency- related matters for the future development of IMT2000 and systems beyond IMT2000 taking into account the results of ITU-R studies in accordance with Resolution 228 (Rev. WRC2003)". Systems beyond IMT2000 are nowadays referred to as IMT-Advanced. IMT will continue to develop and no end date for this enhancement process is currently foreseen. Evolutionary developments of IMT, which impact its technical capabilities, range of available services and breadth of applications, will be progressively introduced during its lifetime. Many regional/industry groups have some preliminary positions on this specific item. For Region 1, the CEPT vision and actions, subject to possible refinements or evolutions, state that: - Globally available frequency bands should be identified for IMT for universal coverage, for 2010 and beyond, to facilitate global roaming and reduce equipment costs through economies of scale;
- Identification of spectrum in the bands below 5 GHz for IMT should be implemented by an appropriate provision in the Radio Regulations at WRC2007;
- The 470-862 MHz band should not be excluded.
Identifying new spectrum bands of less than 5 GHz (the practical upper limit for handheld usage) is extremely difficult, since the only solution is to displace existing users. Some current users of bands below 5 GHz may wish to keep spectrum in these bands, considering that moving to other bands may not satisfy their objectives. They may also request funding to compensate for extra costs resulting from the spectrum re-farming. Therefore the rearrangement of spectrum below 5 GHz will probably happen step by step during at least two World Radio-communications Conferences (WRC) cycles, i.e. 6-8 years, and be subordinated to high-level policy decisions balancing the interests and social benefits of different solutions. ITU-R WP8F is conducting sharing studies in the tentative candidate bands of 450-470 MHz (this band is not considered as a candidate band by CEPT), 470-806/862 MHz, 2300-2400 MHz, 2700-2900 MHz, 3400-4200 MHz and 4400-5000 MHz. Moreover, satellite solutions will also continue to play a key complementary role for terrestrial systems in IMT-Advanced, especially in the context of new "hybrid architectures". Such hybrid solutions, using the same spectrum for both satellite and terrestrial, will also serve to relieve pressure on "terrestrial-only" spectrum. Access to the digital dividend
Definition of the digital dividend
Digital TV broadcasting requires substantially less spectrum than analog broadcast systems - therefore the transition from analogue to digital TV should free a sizeable amount of spectrum in the 470-862 MHz UHF broadcasting band by the end of the transition period. This freed spectrum is referred to as the "digital dividend" [8-9]. However, the exact value of the spectrum released will vary from country to country, and the timings vary across Europe and the world, from 2007 to 2015, as 2015 is the agreed date for the end of the transition period for the UHF band in the GE-06 Agreement of the ITU-R RRC-06 Conference. Candidate applications competing for the digital dividend
Even prior to confirmation of a dividend, there is tremendous market pressure from candidates competing for all or part of these UHF frequencies (Table 3): 
- Broadcasters requiring increased capacity for additional/new digital TV programs, new HDTV channels and local community channels;
- The UMTS Forum, asking for "Rural 3G" mobile allocations of up to 2x40 MHz between 470-610 MHz;
- Mobile TV (also known as Mobile Multimedia Broadcast) providers requesting anything between 470-750 MHz;
- Rural WiMAX: the WiMAX Forum recently launched the concept of WiMAX in sub-1 GHz bands for emerging countries, with Alcatel pushing for the 700 MHz band for WIMAX due to its potential for mass-market development in those countries.
Key regulatory issues Today's UHF frequencies are primarily designated in most countries as broadcast bands. "Broadcast" also includes services such as HDTV, IP TV and even mobile TV, which are treated as natural extensions of the traditional broadcast service (thus having access to broadcast bands). Telecom applications, such as Rural WiMAX or Rural 3G, currently operating under fixed or mobile allocations, would be "new users" of these broadcast bands, requiring additional clarification regarding their regulatory conditions of use. During the last RRC-06 (June 2006), 52 countries supported provisions in the final RRC-06 Agreement to increase the regulatory flexibility allowed by the Digital Broadcasting Plans and the regulatory modification possibilities [10]. Prior to RRC-06, the European Commission encouraged Member States to enable the introduction in the VHF and UHF bands of other services in addition to the existing use of broadcasting services. As pointed out by the Commission, this should be done in two stages: opening the regulatory possibility of using other services during RRC-06; and actually introducing these other services at WRC-07. As a result of the negotiations led by the CEPT countries, the above-mentioned provision in the GE-06 Agreement successfully achieved the first step. Next steps include primary mobile allocation in the 470-862 MHz band in Region 1. In order to take full benefit of the flexibility and to have the possibility to introduce services other than broadcasting, a primary mobile allocation in the 470-862 MHz band in Region 1 at WRC-07 would be required. Such an allocation would: - Resolve the basic issue for the subsequent border co-ordinations, enabling harmonization of non-broadcasting services even in countries that have not signed the declaration;
- Enable global harmonization.
A mobile service primary allocation should be proposed under agenda item 1.4 of WRC-07. Conclusion
Spectrum is an important factor for new business development, since all emerging technologies are spectrum-hungry. It is a limited resource and must be managed efficiently to incorporate new systems, which cannot work without it. The major challenge is to develop a spectrum management framework that supports current radio technology innovations and that should also be commonly applicable - independent of technology/services. Such a framework is needed not only to satisfy the engineering, economic and policy challenges of future spectrum usage, but also to satisfy increasing end-user demand and quality of service requirements. The future spectrum management procedure that considers these three approaches in the long term should be neither a dream nor a nightmare; it should be a reality. This is the only way forward: evolved spectrum allocation and management that satisfy increased user demands while maintaining high-level performance requirements.  | References
[1] Radio Regulations - Edition of 2004 [2] COM (2005)400: A market-based approach to spectrum management in the European Union [3] "The Review 2006 of EU Telecom Rules: Strengthening Competition and Completing the Internal Market", Speech given by EU Commissioner, V. Reding at Annual Meeting of BITCOM, 27 June 2006, Brussels, Belgium. [4] A. Urie, M. Levy, J. Van Bogaert and V. Munière, "Combining Access Technologies to stay best connected", this issue [5] RSPG05-102: RSPG Opinion on Wireless Access Policy for Electronic Communications Services (WAPECS) (A more flexible spectrum management approach) [6] OFCOM analysis provided in the consultation on "Spectrum Usage Rights", 12 April 2006, UK. [7] ECC Report 80: Enhancing harmonization and introducing flexibility in the spectrum regulatory framework [8] COM (2005)204: Accelerating the transition from analogue to digital broadcasting COM (2005)400: A market-based approach to spectrum management in the European Union [9] COM (2005)461: EU spectrum policies priorities for the digital switchover in the context of the Upcoming ITU Regional Radiocommunications Conference 2006 (RRC-06) [10] UMTS Forum input to ECC/PT1 meeting, Oslo, Norway, 26-28 June 2006. |
1 These studies are: "Magic mobile future 2010-2020" (released by the UMTS Forum) and "The demand for future mobile communications markets and services in Europe" (Future Mobile Systems- FMS EC 2005 - available at www.jrc.es), respectively. Both were published in April 2005. A third document, entitled "Development of spectrum requirement forecasts for IMT2000 and systems beyond IMT2000", published by the UMTS Forum, goes even further to calculate the future demand for spectrum from IMT-2000 and systems beyond IMT-2000, up until 2020. 
|
