1. Spectrum OverLay through aggregation of heterogeneous DispERsed Bands Standardisation Plans; Ofcom TV White Spaces Pilot Activities and Associated Standardisation Aspects Oliver Holland, King’s College London Fotis Foukalas, Industrial Sciences Institute Florian Kaltenberger, Eurecom Guillaume Vivier, Sequans Communications Various other SOLDER contributors

2. Introduction to SOLDER 2

3. SOLDER: Bringing together cellular and cognitive radio technologies 3 2G 3G 4G IEEE 1900.6 DSA IEEE 1900.7 ETSI-RRS IEEE 802.22 SOLDER: Cognitive radio application in mobile cellular networks IEEE 802.11af IEEE 1900.1 IEEE DySPAN-SC TVWS

4. Key Aspects of SOLDER Aggregation of heterogeneous bands has not been sufficiently considered in past research, particularly under or involving LTE/LTE- A type technologies SOLDER supports a pervasive radio technology that is able to combine the fragmented spectral and deployed RAT resources akin to a convergence layer Carrier aggregation in heterogeneous networks (HetNets) and radio access technologies (h-RATs) There are many spectrum types considered, including licensed, unlicensed, light-licensed, white spaces and other spectrum: – Fragments of such heterogeneous spectrum must be taken advantage of to efficiently realize future capacity demands – There may be particular requirements and characteristics of deployed systems in those spectrum bands – Aggregation of diverse heterogeneous spectrum opportunities and/or systems therefore necessary 4

5. ETSI-RRS Standardisation Plans 5

6. Standardization Plans – ETSI RRS ETSI RRS WG1: – There is no Work Item that deals with the Carrier Aggregation and to start a new one will not be practical due to the lifetime of the project – The LSA System Requirements document TS 103 154 published in May 2014. Based on this, good to start a discussion on Carrier Aggregation between TDD/FDD systems involving LSA (i.e., LSA TDD and licensed band FDD systems) and provide corresponding requirements – KCL interest in the early draft document EN 303 387 (Signaling Protocols and information exchange, etc.). It would be a worthwhile contribution to ensure that the systems that are going to use TVWS in a coordinated way, and the associated coordination protocol and mechanisms, consider the potential for aggregation of resources by each of those systems – KCL interest in the exchange of information among GLDBs (EN 303 144) that will support better resource coordination decisions in view of aggregation of resources by different systems using different GLDBs 6

7. Standardization Plans – ETSI RRS ETSI RRS WG2: – Contribution related to Software Defined Radio equipment using for example OpenAirInterface of Eurecom. Potential contribution to early draft document EN 103 146-1: Mobile Device Information Models and Protocols: Multiradio Interface – The partner AthenaRC could also contribute to this document from equipment point of view based on their own platform, mainly build by SOLDER. They could also provide contribution to any multi-RAT aspect of the proposal General comment: – In order to provide something efficiently, we have to have very good knowledge of the ETSI documents, e.g., TS, TR, etc., to know where and how could we should contribute – SOLDER is working with CRS-i and others to assist collaboration in these areas and continuous monitoring of ETSI-RRS 7

8. IEEE DySPAN-SC and IEEE 1900 Standardisation Plans 8

9. Standardization Plans – IEEE DySPAN-SC and IEEE 1900 WGs Many opportunities for standardisation in IEEE DySPAN-SC and IEEE 1900 WGs For example, SOLDER members hold and execute responsibilities in a number of leadership positions therein— very strong active participation already in these groups – IEEE DySPAN-SC Treasurer and Leadership Member – IEEE 1900.1 Chair – IEEE 1900.6 Acting Chair – IEEE 1900.7 Vice-Chair – Memberships held of IEEE DySPAN-SC, IEEE 1900.1, IEEE 1900.6, IEEE 1900.7 Of very strong relevance to SOLDER interests 9

10. Standardization Plans – IEEE 1900.1 Scope/purpose of IEEE 1900.1 – Instantiated under the realisation that many of the terms used in the fields of dynamic spectrum access, spectrum management, policy-defined radio, adaptive radio, software- defined radio, reconfigurable radio and networks, and related technologies, do not have precise definitions or have multiple/unclear definitions – Aims to facilitate development of such technologies by clarifying the terminology and aspects of how these technologies relate to each other – Aims to ensure that technologists really understand what each other is talking about when they are collaboratively developing dynamic spectrum access and related technologies 10

11. Standardization Plans – IEEE 1900.1 Envisaged contributions – To ensure that the possible use of spectrum/channel/link, etc., aggregation, in or linked to DySPAN and related technologies, are captured in the terms and definitions of 1900.1 Already one contribution has been brought forward to 1900.1 towards such ends – To ensure that the structures and inter-relationships between concepts in DySPAN and related technologies also reflect the potential and benefits of aggregation of dynamic spectrum access spectrum opportunities, as well as aggregation of those opportunities with licensed, unlicensed and other spectrum usages – Timescale of contribution opportunities overlaps with the entire duration of SOLDER 11

12. Standardization Plans – IEEE 1900.6 Scope/purpose of IEEE 1900.6 – Defines the information exchange between spectrum sensors and their clients in radiocommunication systems – This might be applicable in cooperative/collaborative sensing scenarios, and in other scenarios where the intelligence that makes spectrum access and other decisions and spectrum sensors are at different locations in the network – The logical interface and supporting data structures are defined abstractly without constraining the sensing technology, client design, or data link between sensor and client – Of course, the facilitation of sensing technologies through standards such as IEEE 1900.6 assists many spectral coexistence techniques that utilise locally-obtained spectrum information, such as CR 12

13. Standardization Plans – IEEE 1900.6 Envisaged contributions – Work anticipated to be starting imminently in 1900.6 on the use of 1900.6 to facilitate spectrum databases being augmented by spectrum sensing information – “Spectrum databases” can refer to not only to TV white spaces databases, but also other forms of databases of localised spectrum information – A key issue is knowledge of link/spectrum/channel quality in aggregating link/spectrum/channel opportunities—such knowledge can be supported by spectrum databases – SOLDER can contribute to IEEE 1900.6 through ensuring that the developing work on supporting spectrum databases and associated information exchange between sensors and spectrum databases also covers such possibilities – SOLDER participants have already brought forward past successful contributions on connectivity awareness to IEEE 1900.6, which can be augmented to serve the link/spectrum quality awareness purpose – Timescale of contribution opportunities likely will overlap with the entire remaining duration of SOLDER 13

14. Standardization Plans – IEEE 1900.7 Scope/purpose of IEEE 1900.7 – IEEE 1900.7 is defining a new Radio interface (PHY/MAC) for white space access – The aim is for this radio interface to be generic, applicable to a range of use cases and spectrum bands – However, in practice, it is currently limited to application in TV white space, as TV bands are the only bands currently authorised from a regulatory perspective for white space access 14

15. Standardization Plans – IEEE 1900.7 Envisaged contributions – Contributions anticipated on ensuring that the radio interface developed in 1900.7 facilitates and supports aggregation of spectrum opportunities, both within TV white space, and aggregation of spectrum opportunities in TV white space with other opportunities outside of TV white space – Anticipated contributions on enhancing the FBMC radio interface to assist aggregation of spectrum opportunities – Anticipated contributions to the cognitive plane, assisting aggregation – e.g., smart mechanisms to choose which spectrum opportunities should be used and aggregated – Timescale of contribution opportunities likely to only be throughout 2014, although this may be extended 15

16. 3GPP Standardisation Plans 16

17. Standardization Plans – 3GPP The 3GPP is the body in charge of standardization of LTE, LTE- Advanced and its evolution. As such, it is important for the SOLDER project to stay informed of the evolution of the standardization to steer the technical work aligned with the main stream Consequently, the project monitors on a continuous manner the progress of the 3GPP work, especially with respect to carrier aggregation as, e.g., represented by technical report 36.850 or 36.851. SOLDER is monitoring on a quarterly basis the progress in the RAN (RAN plenary, RAN1, RAN2, RAN4) with sometimes direct attendance to the meeting Similarly, the project attended two specific workshops in January and June 2014 on the topic of the use of LTE in unlicensed spectrum 17

18. Standardization Plans – 3GPP Solder position is to monitor mostly CA and LTE-U; contribute if possible (consortium not dimensioned to impact 3GPP). Some topics that the project will strive to contribute to, if the opportunity arises are as follows – Carrier aggregation Carrier aggregation is a feature introduced in Rel.10 Most of mechanisms are well defined However, Rel12 and further introduces enhancement e.g. – Much more frequency band combinations – FDD + TDD – LTE-U or LAA LTE-Unlicensed or License Assisted Access A new study item for Rel. 13, (not yet approved, but with good traction) To operate LTE in unlicensed spectrum (typically 5GHz) and ensure fair access of all technology Most likely based on carrier aggregation with Primary Carrier on licensed spectrum and secondary carrier in the unlicensed band – Dual connectivity A way to handle two connections – LTE + LTE – LTE + other technology 18

19. Ofcom TV White Spaces Pilot Activities and Associated Standardisation Aspects

20. Our Trial within the Ofcom TV White Spaces Pilot Almost unique as an academic-led participation in the Ofcom Pilot One of the most extensive and varied (perhaps the most extensive and varied) trials in the entire Ofcom Pilot Leading the involvement of numerous high-profile projects and organisations with interests in spectrum sharing, TV white spaces, opportunistic spectrum usage, cognitive radio, etc.

21. Our Trial within the Ofcom TV White Spaces Pilot - Objectives To test communications systems and scenarios that may be implemented in TV White Space – LTE multicast/broadcast (eMBMS) – Broadband for public protection and disaster relief – TD-LTE and other TDD systems for more general applications in TV White Space (e.g., general broadband provisioning, and small cells in TV White Space) – WiFi in TV White Space (802.11af draft) – Wireless backhaul links in TV White Space – M2M implementations (possible future work) To support the development/assessment of the ETSI 301598 standard

22. Our Trial within the Ofcom TV White Spaces Pilot - Objectives To test the correct performance of the UK’s TV White Spaces framework in general To carry out research studies using TV White Space implementations – Aggregation of resources/links (e.g., TV White Space with licensed and other unlicensed such as ISM, and links within TV White Space) Qualitative and quantitative performance assessments Development also of protocols supporting aggregation – Secondary coexistence (e.g., LTE coexisting and 802.11af in TV White Space, and multiple instances of different standards/devices coexisting) – To undertake studies and surveys on the performances that are achieved, e.g., in terms of interference to primary (!), secondary user performance through objective user opinion polling – Spectrum monitoring and assessment (e.g., spatial and temporal effects on the spectrum—correlation)

23. ETSI 301 598 Harmonised European Standard Defines the technical requirements to avoid harmful interference – RF parameters, e.g., Transmission bandwidth and spectrum mask class Maximum RF power (and compliance thereof) Unwanted emissions (out of TV band) Transmitter reverse intermodulation – Logical specifications, e.g., Control and monitoring Geolocation capability Software, firmware and user access restrictions Geolocation database discovery Data exchange with geolocation databases Defines the testing procedures for ensuring that those technical parameters are conformed with Device to geolocation database communications protocol not defined

24. ETSI 301 598 Support Our trial is undertaking extensive work related to ETSI 301598 Assisting conformance assessment of devices for certain entities Supporting assessment of its effectiveness; providing feedback on ETSI 301 598—links with ETSI

25. Bottom Line Facilitating the realisation of TV white spaces and aggregation of spectrum opportunities thereof, and also undertaking an extensive series of research experiments in the pilot on opportunities for spectrum/link aggregation in or involving TV white spaces Extensively supports the development and testing of the ETSI 301 598 standard, which specifies the conformance requirements for white space devices and certification testing procedures for those devices Also aimed to use the work to provide feedback and further refinements/inputs to ETSI 301 598

26. Conclusion 26

27. Conclusion Several standardisation involvements and opportunities, within ETSI-RRS, IEEE DySPAN-SC and IEEE 1900, and possibly 3GPP (monitoring opportunities for the latter) Some activities in this regard have already started Leading of an extensive trial within the Ofcom TV White Spaces Pilot, with many aggregation aspects therein Very strong testing and feedback/inputs to harmonised European standard on conformance and certification for white space devices (ETSI 301 598) Open to discussion with those that want to participate in our Trials within the Ofcom TV White Space Pilot, and standardisation assessment/feedback/input elements thereof (ETSI 301 598) 27

28. 28 Thank you! oliver.holland@kcl.ac.uk

29. Acknowledgement This work is supported by the ICT-SOLDER project, FP7 project number 619687, www.ict-solder.eu, and the ICT- ACROPOLIS Network of Excellence, FP7 project number 257626, www.ict-acropolis.euwww.ict-solder.euwww.ict-acropolis.eu 29

30. Back-up Slides 30

31. LTE MBMS and opportunistic spectrum/link aggregation with other services (WiFi in ISM, and 3G/4G in licensed bands) Augmented broadcast (e.g., extra layers of video subscribed to when receiving higher rate, locally customised broadcast) Data carousel-like functionalities achieved by raptor-coding the data set Augmented CPC Software upgrades Deployment and Testing Scenarios: LTE MBMS and Spectrum/Link Aggregation

32. LTE femtocells + intercellular links in TV White Space Quickly-deployable field solutions for emergency situations (e.g., enhanced provisioning or coverage extension to emergency workers) Ad-hoc repair of communications links (e.g., backhaul) in disaster scenarios (e.g., earthquakes) Deployment and Testing Scenarios: Public Protection and Disaster Relief

33. Video surveillance system in TV White Space Deployment and Testing Scenarios: Public Protection and Disaster Relief 2 Sony SNC-CH220 + 1 Carlson Terminal 1 Sony SNC-ER550 + 1 Carlson Terminal Sony Real Shot manager software Carlson Basestation

34. Point-to-point links for backhaul provisioning, between different university campuses of participants in our trials (one challenging example we will attempt to achieve is to the right) General broadband provisioning using a range of devices and systems LTE small cell implementations Wireless local area networking in TV White Space Machine-to-Machine communications in TV White Space (possible at later stage) Deployment and Testing Scenarios: Others Denmark Hill Mile End ~7 km

35. Devices: Eurecom ExpressMIMO2 ExpressMIMO2 is the basis for the LTE MBMS case initially, and likely other LTE cases later—perhaps also 802.11af at a very late stage No DSP on board, FPGA primarily used just for routing data; host PC must be powerful and running in a real-time operating system!!! 4 RF chains achievable on the card (all Tx+Rx) Have set up 3 devices based on this so far (1 base station and two terminals), each hosted in a PC with (in the case of base station) a separate box handling RF

36. Devices: Eurecom ExpressMIMO2 Perhaps the first Class 1 white space device implementation? Has been a significant challenge in achieving this—Class 1 constraints are very tough ExpressMIMO2 has excellent RF performance, but on some channels it fails Class 1 marginally for the adjacent channels only, reducing to Class 3 To still achieve Class 1, signal has been created at a high fixed frequency and very precisely filtered there Down-converted with a variable frequency LO, to allow switching to the different TV channels A complex filtering solution is employed to get rid of all images and other issues (e.g., imperfection of the mixer leading to some output remaining at the high signal frequency). The high signal frequency has been carefully chosen to ensure that such issues are only outside of the TV bands so can be very precisely filtered with fixed filters Amplifier with extremely large back-off (hence linearity) used to ensure that Class 1 performance is maintained at amplification

37. Devices: Carlson Wireless Ruralconnect http://www.carlsonwireless.com/ruralconn ect http://www.carlsonwireless.com/ruralconn ect Built for US market, but adapted to operate under Ofcom/ETSI rules in terms of database (and database of databases) communication, power levels, etc. Our trial will use at least 2 base stations and 5 terminals Deployment scenarios include the public protection and disaster relief cases Also broadband provisioning cases, and perhaps to try to use for longer-distance point-to-point links at a later stage

38. Sinecom/KTS Agility White Space Radio http://sinecom.net/product.ht ml http://sinecom.net/product.ht ml In the shorter term, to be used for low-rate broadband provisioning In the longer term, likely to also be used for M2M cases Likely to be used for the point- to-point long-distance links at a later stage Our trials will have at least 6 of these devices

39. Devices: NICT Devices (collaboration with NICT) TD-LTE in TV White Space – Base stations and terminals – 3 of each will be used in our trials – Used for general testing of LTE scenarios (small/femto cells, and larger cellular provisioning cases) Low-power IEEE 802.11af (WiFi in TV White Space) – Wireless local area networking is prime use case – We will have at least 3 of these devices High-power IEEE 802.11af – Long-distance backhaul link provisioning – We will have at least 2 of these devices

40. Devices: NICT Devices (collaboration with NICT) Wireless mesh network deployment example at NICT, Yokohama, Japan (very low Tx power in this case), also with graphical representation of the NICT database implementation

41. Databases Noted that the interfaces between TV White Space devices and geolocation databases are not standardised. It is therefore typically the case that given TV White Space device manufacturers are working with particular databases We are using a range of databases in our trials – Fairspectrum  Carlson Wireless and Eurecom devices – NICT  NICT and Eurecom devices – Spectrum Bridge  KTS/Sinecom devices – Joint Research Centre of the European Commission  for comparison using a range of devices, not deployed in UK Haven’t pursued the implementation details yet simply due to time constraints, but also have been in discussion and have verbal agreement with the following – hope to test devices with these databases too as trials progress – BT – Sony – Nominet

42. Locations  Almost exclusively campuses/buildings among the range of universities that are collaborating in our trials  Extensive range of locations, covering almost all imaginable environments, tested (mostly) sequentially - Cluttered vs. non-cluttered - High incumbent systems TV bands usage vs. relatively low usage - A range of propagation characteristics

43. Locations London – King’s College London Denmark Hill – King’s College London Guys (London Bridge) – King’s College London St. Thomas’ (opposite Westminster) – King’s College London Hampstead – Queen Mary University of London – King’s College London Strand – long-term objective, dependent on whether coexistence challenges can be managed – King’s College London Waterloo – long-term objective, dependent on whether coexistence challenges can be managed Outside London – University of Surrey (Guildford) – University of York – Strathclyde University (Glasgow—under discussion) – Cambridge University – University of Bath – Leeds University (back-up)

44. Research Examples - Aggregation Solutions for Aggregation of resources/links (TVWS resources aggregated with licensed and unlicensed ISM, and channels aggregated in TVWS) – As well as assessing performances, to look at technical means of achieving aggregation compatible with ETSI/Ofcom rules (e.g., link bonding at higher layers, cross-band scheduling decisions, etc.) – LTE in unlicensed spectrum (LTE-U) one among many interesting cases – Why not such a LTE-U supplemental downlink in TV White Space license- exempt spectrum opportunities? Qualcomm White Paper, “Extending LTE Advanced to Unlicensed Spectrum,” December 2013

45. Research Examples – Primary Service Coexistence Assessment Dedicated equipment to look at effect on DTT, e.g., Wavecom devices – Signal Power, Modulation Error Rate, SINR, CINR, BER before Viterbi, BER after Viterbi, BER after Reed-Solomon, etc. Will devise challenging scenarios to interfere with DTT, within the scope of ETSI/Ofcom rules (e.g., indoor TV antennas in same room as white space device, saturating TV antenna amplifiers, etc.) Also plan to test interference with PMSE through our own PMSE equipment, again within Ofcom/ETSI rules. E.g., blind online surveys

46. Research Examples – Spectrum Monitoring and Statistical Inferences Long-term fixed measurements or spatially distributed measurements, to assess the effects on the spectrum of TV White Space devices – Assessment of correlation aspects of spectrum usage both with and without white space devices present (useful for, e.g., assessing the spatial uncertainty in the effects on the spectrum) that white space devices may have – One monitoring location on roof of King’s College London Strand Campus

47. The Trials Team ACROPOLIS Project – Led by King’s College London, UK The Joint Research Centre of the European Commission, EU Eurecom, France – Also involving RWTH Aachen University, Germany Saints’ Cyril and Methodius University in Skopje, FYRoM Poznan University of Technology, Poland University of Rome “La Sapienza”, Italy University of Piraeus Research Centre, Greece Institute of Accelerating Systems and Applications, Greece University of Surrey, UK University of Leeds, UK

48. The Trials Team Extensive involvement of other projects, notably ICT-SOLDER (www.ict-solder.eu), ICT-CREW (www.crew-project.eu), Newcom# Network of Excellence (www.newcom-project.eu), ICT-CRS-i (http://www.ict-crsi.eu). Also numerous high-profile individual groups participatingwww.ict-solder.euwww.crew-project.euwww.newcom-project.euhttp://www.ict-crsi.eu Following reflects both the above projects participants, and individual groups participating (not exhaustive) – Belgium: iMinds, IMEC – Finland: Fairspectrum, Turku University of Applied Sciences – Germany: Technical University of Dresden – Greece: Industrial Sciences Institute – Ireland: Trinity College Dublin – Italy: CNIT/Politecnic of Torino, Fondazione Ugo Bordoni, Create-Net – Japan: NICT, Sony – Portugal: IT/University of Aveiro, IT/University of Beira Interior – Slovenia: Jozef Stefan Institute – UK: Queen Mary University of London, University of York, University of Cambridge, University of Bath, University of Strathclyde/Larkhill, British Telecom, Nominet