1. Existing & Near Future Work in CEPT on Cognitive Radio 56 th International Symposium ELMAR Zadar, Croatia, 10-12 September 2014 Thomas Weber, Spectrum Management, European Communications Office

2. Content TV WSD (White Space Devices) – Activities in the ECC Other ideas with regard to geo-location database usage Cognitive PMSE Tendencies for SRD applications’ use of mitigation techniques 5 GHz considerations with regard to mitigation techniques Licensed Shared Access (LSA) Sharing principles for applications under general authorisations Broadband Applications – Bundling of applications under the same technical framework What does it mean regarding spectrum occupancy measurements (SOM) / monitoring – how can this help? Conclusions - Outlook

3. TV WSD (White Space Devices) – Activities in the ECC What we have in CEPT: ECC Report 159: Technical and operational requirements for the possible operation of cognitive radio systems in the ‘white spaces’ of the frequency band 470-790 MHz ECC Reports 185 “Further definition of technical and operational requirements for the operation of white space devices in the band 470-790 MHz” ECC Report 186 “Technical and operational requirements for the operation of white space devices under geo-location approach”. WGFM: A summary and analysis of the answers to the questionnaire on PMSE protection in case of introduction of WSD in the 470-790 MHz band. In progress: WGFM PT FM53 work item for creation of a new ECC Report for TV-WSD (White Space Devices) using geo-location. The aim is to describe the overall regulatory framework for TV WSD using geo-location databases and guidance for national implementation. Delivery of this new ECC Report not before mid 2015. Very limited contributions so far. Only afterwards, possible creation of a harmonisation measure (e.g. ECC Recommendation). ECC never received a request (SRdoc) from ETSI on TVWSD.

4. Who uses all or part of the 470-862 MHz range for PMSE? 26 countries out of 27 questionnaire responders use for PMSE and will continue using it: 10 provide individual licenses; 7 provide general authorisations; 9 both. 13 countries consider changes for PMSE authorisarions Do you envisage the introduction of white space devices in the 470-790 MHz band? PMSE includes inter-alia Microphones and in-ear monitor systems (PWMS), Stage direction radio technology (Wireless Intercom), Cameras (equipped with the necessary audio and video receiver and transmitter), Effects and remote control technology, Conference systems, Transmission systems for translated conference contents, (Tour-) Guide systems

5. Is there a need for a specific mechanism? 13 countries said YES (ONLY 50% of all answers) Safe harbour/ geolocation most mentioned, also as combination. Safe harbour seen as simple and providing higher protection for PMSE, however also as inefficient and freezing spectrum; DBs are more efficient than manual licensing / coordination but only 7 countries think there is sufficent information available; Safe harbours: 5 regional, local and 4 nationwide

6. Industry Side Do you envisage requesting your Administration or CEPT the introduction of white space devices in the 470-790 MHz band? Only one respondent (out of 14) answered with YES (answers received mainly from Broadcasting and PMSE community); Comparison: at the same time, ECC has received requests in a total of 6 (!) ETSI System Reference Documents for future spectrum on a variety of application in the UHF band for other frequency bands (SRD, Automotive, Smart Metering, Smart Grids, RFID, Alarms, Home Automation etc); No trust in future TV WS solution yet. FCC: incentive auctions

7. Industry and User Responses – Which mechanisms? (I) Do you consider specific mechanisms are necessary? 13 answers: the online registration/ database and the safe harbour concept for PMSE were indicated the most and as a preference; Except for one answer, cognitive solutions are seen critical if not unfeasible or not mature enough. 3 answers indicated that a database concept might be beneficial for large events -> this is also an ACCEPTANCE PROBLEM; While some indicate that all categories of PMSE should be registered, 2 indicate to differentiate between spontaneous and predictable usage, 2 others indicate that registration is only needed under certain circumstances (e.g. large events). One answer said it was not possible if applications are license exempt. One answer recommended that this needed to be defined together with the user groups. Note: Japan introducing a geolocation concept with full registration/coordination

8. Industry and User Responses – Which mechanisms? (II) Safe harbours for PMSE: the answers vary: starting from 32 or 60 audio channels to 2, 3, 2-4, 4-5, 4-6, 2*3 or even 10 TV channels. Some indicated that the number of channels is unknown (4) or there would be no place for such channels (1). Need for harmonisation: 2 indicated local needs, whereas most indicated nationwide/ preferably nationwide channels, 4 indicated the preference for harmonisation on European level. Cognitive, beacons: The answers provided indicate that although the potential of beacons or cognitive approaches is recognised, the techniques are not considered mature enough, efficient enough or to solve all problems.

9. Cognitive PMSE (C-PMSE) New draft ECC Report 204: Spectrum Use and Future Requirements for PMSE (in public consultation) Initial research activities on cognitive PMSE systems have been initiated in ETSI STF 386 and in a German research project funded by BMWI (German Federal Ministry of Economics and Technology) called C-PMSE. Great Potential: Features several in-operation audio PMSE links can be transferred to less impaired spectrum when necessary – w/o noticeable interruptions Can lead to more robust and more efficient use of spectrum, especially in UHF bands, including duplex band operations It’s cognitive because the used frequencies are ”monitored” constantly / service quality checked all the time. Existing audio PMSE systems check only before tyhe event A reaction to the changes in the UHF terrestrial broadcast spectrum ETSI Technical Specifications as well as first demonstrator hardware available It’s monitored in the ECO Research Activities w.r.t impact on work in the ECC

10. Other ideas with regard to geo-location database usage FM PT44 (satellite services) received a presentation on the CoRaSat project, which looks at different sharing scenarios between satellites and other terrestrial services (mostly Fixed Service) or other satellite applications, using cognitive and database methods. ETSI TC SES is creating an ETSI SRdoc (avail. expected mid 2014). FM PT44 will deal with this. E.g. improved FSS Reception at locations “outside” FS Existing work items in WGFM/SRDMG and WGSE/SE24: Some fixed installed Applications such as Assistive Listening Devices in the VHF band (174-216 MHz) or in other spectrum (e.g. for telecoil replacement systems) are discussing geolocation approaches. However: geolocation is only one of the options for sharing under discussion, other options such as Sensing LBT/frequency agility or bundling applications in spectrum which are unlikely to be operated at the same location may be more efficient (examples: ALDs and RFID reader channels in the 915-921 MHz, uncoordinated FSS in FS duplex gaps) and involve less burden in terms of technical complexity or for the administrations.

11. Tendencies for SRD applications’ use of mitigation techniques Example: New band plan 870-876 MHz (from draft ECC Report 189; currently in public consultation – work triggered by 5 ETSI Srdocs – plus 6 th Srdoc for WB apps) Also: ECC Report 182 (Survey about the use of the frequency band 863-870 MHz) Duty Cycle Mitigation is very dominant – Sensing not part of it !

12. 5 GHz considerations with regard to mitigation techniques The draft ECC Report 192 on the current status of DFS (Dynamic Frequency Selection) in the 5 GHz frequency range was in public consultation. Severe interference into weather radars (> 200 reported cases in 2012) in a considerable number of European countries caused by outdoor LS-links links using WAS/RLAN equipment. Reasons: mainly intentional illegal use or non-compliant equipment. There is some motivation to switch off the DFS! Mitigation must fit to the application/intention (Bad examples: UHF RFID with LBT, Direct-Air-to-Ground with DAA to protect BFWA) Recent studies on WIA in SRD/MG and PT SE24 (Wireless Industrial Applications with higher emission levels up to 400 mW in 5725-5875 MHz: DFS (protection of radiolocation), DAA (protection of BFWA, ITS, road tolling) – ALWAYS -, APC based on potential outdoor use (indoor use is not the problem). 5725-5875 MHz≤ 400 mW e.i.r.p.APC required Adequate spectrum sharing mechanisms (e.g. DFS and DAA) shall be implemented by the equipment. Modulation bandwidth ≥ 1 MHz and ≤ 20 MHz Wireless Industrial Automation (WIA) Networks The Adaptive Power Control has a range of at least 12 dB (dynamic range between 400 mW and 25 mW)

13. Licensed Shared Access (LSA) The draft ECC Report 205 on Licensed Shared Access has been approval for public consultation – it includes some refinements of the LSA concept as set out in the respective draft RSPG opinion. Scope: LSA is a complementary spectrum management tool that fits under an “individual licensing regime”. LSA aim to facilitate the introduction in a frequency band of new users, which require a certain level of guarantee in terms of spectrum access, while maintaining incumbent services in the band. LSA licensees and incumbents operate different applications and are subject to different regulatory constraints. They would each have exclusive individual access to a portion of spectrum at a given location and time.

14. LSA (II) The first practical use case of LSA will be to provide access to additional spectrum for mobile broadband services (MFCN) in 2.3-2.4 GHz. The implementation of LSA relies on the concept of a “sharing framework” that is under the responsibility of Administration/NRA. Its development requires the involvement of all relevant stakeholders. The “sharing framework” can be understood as a set of sharing rules or sharing conditions that will materialise the change in the spectrum rights of the incumbent(s) and define the spectrum, with corresponding technical and operational conditions, that can be made available for alternative usage under LSA. The practical implementation of LSA as a complementary solution for mobile network operators for accessing spectrum for MFCN in specific bands, within specified geographical or technical limits and on the requirements to be included in CEPT “harmonisation measure”. The first harmonisation measure developed by CEPT that includes the possibility for implementation of LSA is for the introduction of MFCN in 2.3-2.4 GHz band. It can include geolocation !

15. Sharing principles for applications under general authorisations The SRD strategy and sharing principles are set out in CEPT Reports 14 and 44. 1.Equal access to spectrum (and incumbents should not enjoy “seniority”) 2.Technology- neutrality 3.Application- neutrality (as much as possible) 4.Investigate existing SRD spectrum and existing SRD categories first before considering new spectrum 5.Avoid fragmentation of the spectrum, search for more efficient sharing solutions 6.Foster innovation (broad SRD regulation is more open to new ideas than specific regulations) Examples: 1.New band plans 870-876, 915-921 MHz are more horizontal than vertical (e.g. 863-870 MHz) 2.Medical Applications: LP-AMI, MBANS, Cochlear in 2483.5-2500 MHz are foreseen to share on equal basis and using spectrum access mechanism such as LBT with frequency agility)

16. Band Plan 915-921 MHz Note: RFID/ALD sharing Safe harbours Flat SRD regulation RFID DAA for E- GSM-R

17. SRDs: What you often have to measure.. (data comes from surveys, not from measurements) Techniques Bandwidth Application Max Cumulated TxON timeMax equivalent DC over 1 second [in seconds] with current definition (per 1 h) Automotive100ms remote keyless entry0.150.002% convertible roof10.139% TPMS0.030.001% ITS CAM0.0150.125% Home & building control10ms Mains powered devices0.0250.003% 0.20.005% Battery powered devices0.60.007% 10.012% Repeaters0.0250.001% Smoke detectors10.00139% Low cost point to point devices10.012% Telemetry, telecomand350ms to 1s 11.620% Metering25ms to 1s without in home display0.0250.000029% with in home display0.0250.003% Repeaters0.0250.001% EN13753 Mode R210.023% Alarms25ms to 1s Intrusion alarm0.0250.001% Social alarm0.150.001% Battery power devices10.028% Imaging10.007% Referee voice system0.10.417%

18. Broadband Applications – Bundling of applications under the same technical framework Example: 2 GHz Unpaired Bands: as a result of the synergies between PMSE and PPDR, as well as between DECT and SRD, the shortlist indicated in the EC Mandate, composed of five potential harmonised uses of the 1900-1920 MHz and 2010-2025 MHz frequency bands (DA2GC, PMSE, PPDR, DECT and SRD), is considered to be adequately studied by using 3 usage blocks. Video PMSE / ad-hoc PPDR: coordinated applications – QoS requirements quite high – advanced coordination is already difficult. Advanced monitoring in the ”unplanned event” case not possible. BB-Applications in general: we need to bundle, otherwise we run out of frequencies ! DECT/SRD: DECT proponents already recognised need for technology-neutral regulation. SRD again may use low duty cycle mitigation. DECT uses DCS. No motivation for further mitigation mechanism developments. BB-Direct air to ground communications DA2GC Video links and cordless cameras PMSEPPDR Applications under general authorisation DECTSRD

19. What does it mean regarding spectrum occupancy measurements (SOM) / monitoring – how can this help? All sharing situations are pretty specific – including all the technical parameters – a general SOM approach cannot help. Customisation is needed. In most situations, it will be nearly impossible to measure the real occupancy of the spectrum due to the incumbent service/application characteristics (e.g. antenna characteristics, low duty cycle, low emissions, temporary etc) – The fixed SOM approach is not possible in almost all UHF situations. There are often severe acceptance problems (“interpretation of data”) regarding cognitive sensing approaches and the same is valid for spectrum occupancy measurements with regard to related spectrum management questions Very often, alternative mitigations without sensing may be more effective / less complex Change of spectrum occupancy can create severe problems (pre-emption not tolerated, spectrum access must be predictable to a certain extent) – the application expectations do often not fit to a spectrum management based on sensing/ SOM Complimentary information (i.e. availability and completeness) from SOM is not sufficient to design for improved co-existence. Sharing concepts would rather “jump” from low complexity to real co-existence standards.

20. Lessons learned – 5GHz DFS mechanism DFS mechanism -> sensing radar patterns for the protection of the radiolocation service (military as well as weather radar). The ECC Report 192 on the current status of DFS (Dynamic Frequency Selection) in the 5 GHz frequency range was published in 2013. Severe interference into weather radars (over 200 reported cases in 2012 in the CEPT) in a considerable number of European countries caused by outdoor fixed service links using WAS/RLAN equipment. The reasons were mainly intentional illegal use or non-compliant equipment -> there was some motivation to switch off the DFS!

21. Conclusions Cognitive techniques need to be analyzed not only from the perspective of the technical feasibility but also from the perspectives of the demand, fit to application, investment requirements, the level of technical complexity involved, and the long term predictability of the spectrum access, this may include intra- SRD /WSD compatibility needed (minimum predictable spectrum access!) New CEPT Report in response to the 6th update process of the European Regulations on SRDs may rather present the initial results of these investigations and not include a proposal for a final regulatory approach/implementation but it is intended to make a step forward and may lead to an implementation proposal in a second step. CEPT is tasked with regularly updating of the technical Annex of EC Decision 2006/771/EC. The guidance letter to the CEPT includes the request to study cognitive radio enabled SRDs. -> Questionnaire -> to investigate the above points

22. How research projects can feed the existing processes Almost all examples in this presentation used the ETSI- ECC coordination mechanisms (many SRD ETSI system reference documents, also LSA, TC SES, DECT aso.) Relevant principles are already in place in ETSI when preparing a new ETSI System Reference Document for the designation of spectrum for new demand. The SRdoc has to include the necessary justifications (e.g. information, why the application cannot fit into existing regulation or needs dedicated spectrum,, sharing possibilities) and undergoes a consultation process throughout the complete ETSI membership that gives often also existing spectrum users the possibility to include their view in the document. CEPT is subsequently provided with a consolidated ETSI view and not single interests. See: http://www.cept.org/ecc/who-we-are/ecc-and-etsihttp://www.cept.org/ecc/who-we-are/ecc-and-etsi Another possibility is to approach ECC Project Teams and Working Groups and provide a presentation about your project. It can also be linked to the ECO Research activity (so we monitor w.r.t. impact/link on work in the ECC!)

23. Thank you for you attention Thomas.Weber@eco.cept.org www.cept.org/eco www.cept.org/ecc Questions??