OFCOM SPECTRUM ADVISORY BOARD HELD ON 16 MAY 2016 AT RIVERSIDE HOUSE, 2a SOUTHWARK BRIDGE ROAD, LONDON SE1 9HA FROM 1.00PM

Transcription

1 OFCOM SPECTRUM ADVISORY BOARD HELD ON 16 MAY 2016 AT RIVERSIDE HOUSE, 2a SOUTHWARK BRIDGE ROAD, LONDON SE1 9HA FROM 1.00PM Present: Apologies: David Meyer (Chairman) Greg Bensberg Mischa Dohler Linda Doyle David Harrison David Hendon Nick Munn Niall Murphy Jean-Jacques Sahel Mike Walker Gavin Young Other Ofcom Colleagues Phillipa Marks and Robert Pepper. Action Item 1: Minutes of last meeting held on 26 October 2015 OSAB 52(15) 1. The minutes of the meeting held on 26 October 2015 previously circulated to the Board were agreed as an accurate record for signature by the Chairman. Item 2: Chairman s Introduction and Matters Arising 2. The Chairman welcomed everyone to the fifty-third meeting of the Committee. He thanked Nick Munn for being able to attend the meeting in place of Jon Fuller. The Chairman advised the members that, following a job change, Robert Pepper had decided to step down from the Committee as he was no longer able to attend meetings. Item 3: Access Network Frequency Plan 3. The Chairman introduced representatives from BT who would deliver an update on BT s Access Network Frequency Plan (ANFP). The members noted the following points: NICC DSL Task Group specifications and guidelines relating to the use of Digital Subscriber Line technology in the unbundled UK metallic access network (BT, KCOM) Purpose of the ANFP o To ensure harmonious coexistence of different types of DSL technologies that operate in the shared metallic access network where signals can couple from one pair to another

2 (known as crosstalk) o To set differential mode power spectral density limits to allow Communication Providers to deploy and manage their own xdsl, and manage crosstalk o To be used as an agreed UK standard contractually when Communication Providers take Metallic path facilities (MPF) from Openreach G.fast was originally designed to be the DSL technology that enables FTTDP to deliver Gigabit/s speeds over the D-side network G.fast can be both a FTTDP and FTTC copper DSL technology to deliver Gigabit/s speeds over the D-side network The signal spectrum used o G.fast uses frequencies up to 106MHz (and possibly up to 212MHz in future) o The power spectral density is defined in ITU-T G.9700 and the transceiver in G.9701 o Spectral compatibility with other technologies is key o Flexibility to shape and notch the G.fast spectrum is provided o The NICC DSL TG used these two core ITU-T specifications to underpin the work on defining the spectrum for G.fast in the UK via the BT ANFP (ND1602) G.fast spectrum (ITU-T) o Differential mode PSD limits are defined in G.9700 o The limits exist for EMC purposes and to limit NEXT and FEXT in the multi-pair access environment o These limits were approved by the ITU-T in co-operation with ITU-R o Early technology will use the 106MHz profile What s new in the BT ANFP (ND1602) o Two new parts (E and F) added to cover use of G.fast technology in the D-side network o These parts define the power spectral density (PSD) of the signals injected into the access segment o They cover the use of Time Division Duplexing technology o Part E covers the downstream transmission (eg PCP or Cabinet to customer premise) o Part F covers the upstream transmission (eg customer premise to PCP) o Part F is tied to Part E o The updated BT ANFP was published in March 2016 by NICC after a year s study in the DSL Task Group to incorporate these changes Spectrum compatibility approach o G.fast can operate over the frequency spectrum from 2MHz to 106MHz o As this overlaps with the VDSL2 spectrum (0.14 to 17.7MHz) it is essential to preserve spectrum compatibility to avoid interference between VDSL2 and G.fast signals

3 operating in the same cable binders (this can be achieved by starting the G.fast spectrum at ~19MHz, well above the upper frequency of VDSL2) o More complex techniques will develop to allow spectrum overlap and increased G.fast performance. NICC DSL TG is beginning the study of these Key G.fast standards o ITU-T Rec G.9700 Fast access to subscriber terminals (G.fast) Power spectral density specification Rec G.9701 Fast access to Subscriber Terminal (FAST) Physical layer specification Rec G Physical layer management for G.fast transceivers o Broadband Forum TR-301 Architecture and Requirements for Fibre to the Distribution Point ID-337 Fast Certification Test Plan (member only) TR-285 Copper Transmission Models for testing above 30MHz o NICC DSL TG ND1602 Specification of the Access Network Frequency Plan applicable to transmission systems connected to the BT Access Network 4. The members discussed the presentation and offered the following comments: That performance might be effected by interference from noncompliant equipment used knowingly or unknowingly by consumers. Whilst this was really a matter for Trading Standards there could also be an issue for Ofcom with respect to interference investigation. Whilst G.fast would transmit direct to the home it was not clear as to how a consumer would raise a complaint. Whether consumers would need special equipment to receive G.fast (the BT representatives said that a G.fast enabled modem would be made available and that a splitter filter would also be required) That there could be problems with the quality of the BT signal at the termination point (the BT representatives acknowledged that some problems had been encountered and fixed. It was noted that problems could arise as a result of the age of the network as well issues arising from consumers self installing. BT would be able to offer the services of an engineer at a cost to the consumer to resolve any installation problems arising from self install). There was the possibility of an impact on a broadband universal service obligation for broadband if G.fast moved to a lower frequency (the BT representatives noted that this could be addressed by extending the VDSL frequency to lower band thereby making a lower noise floor)

4 That the balance of usage of G.fast would be in urban areas but it was not necessarily restricted to such areas (OSAB noted that trials were underway) Whether there would be issues with FM radio (the BT representatives did not believe that issues with FM would arise; nevertheless they would have the capability to investigate such matters) That investment in G.fast could be at the expense of investment in fibre (the BR representatives thought that this would not be the case they foresaw a need for both G.fast and fibre and noted that currently there were 10 million homes that could not be supplied via fibre Item 4: Broadband universal service obligation 5. OSAB received a presentation on the subject of a universal service obligation for broadband. The members noted the following points: Ofcom was advising the Government on the design of the broadband universal service obligation (USO) o The Government had proposed a broadband USO with a download speed of 10Mbit/s The Government saw USO as a safety net allowing consumers to access online services Ofcom argued in 2013 that 10Mbit/s was needed given a typical household s use of digital services Ofcom had reiterated this view since then eg in the Strategic Review of Digital Communications (DCR) o In 2015, 2.4m premises could not receive download speeds greater than 10Mbit/s This equates to over 8% of premises Around 1.5m (48%) of rural premises were affected. Scotland, Wales and Northern Ireland were disproportionately affected are they are more rural than the UK as a whole A significant number of SMEs were affected A combination of BDUK superfast rollout and continued private investment would reduce this number but it would remain a significant proportion across the UK o Ofcom was consulting stakeholders on the design of the USO DCMS has requested technical advice and recommendations on the design of the USO from Ofcom Advice is requested by the end of 2016 Ofcom has published a call for inputs (CFI) seeking input from industry and consumers on the design of the USO. The consultation would close on 23 June.

5 o Main areas covered in the consultation included What should the technical specifications be? (ie should upload speeds, jitter, latency and contention, as well as download speeds be specified) Was pricing intervention needed to ensure that USO was affordable? (ie should uniform pricings or caps on charges be set. Should there be a social tariff for consumers on low incomes or with special social needs) What might demand be? (the number and location of USO premises will affect the overall cost of the policy. Take up will also affect the per premises cost of the policy) How much will the USO cost? (overall cost would depend on the technology used to deliver the USO and the scale and location of demand. Both the overall cost of the policy and the cost of upgrading any individual connection must be efficient and proportionate) Who would be the universal service provider (USP) Ofcom could designate the USP directly or by a competitive process. There was a preference for a competitive process but this would depend on whether there were sufficient credible bidders to be the USP How will the USO be paid for? The Government s preference was for an industry-funded mechanism. Would the USO cause market distortions? Minimising the risk of existing network overbuild would retain incentives for operators to invest in highspeed networks. Ofcom would also want to minimise the impact to retail competition. When should the USO be reviewed? The performance delivered by the USO will need to increase over time to continue to meet users needs. Any timeline for review will need to consider the investment lifecycle of the USP s network investment. Ofcom was considering whether to set a minimum period for review 6. The members discussed the presentation and offered the following comments: That not spots and problems with provision in rural areas would need to be identified and solutions devised (such as upgrading the network) What would be the timescale for the delivery of 10Mbit/s. If it was two years it would be challenging but not if it was five years. OSAB noted that this was a manifesto commitment and, as the Government would be judged on its performance, it would need to

6 be delivered by 2019/20. Quality of service issues would need to be included in the USO (eg such as access to emergency services) Ofcom would need to understand the impact of reductions in speed at peak times. Measuring this would not be a trivial exercise. In terms of service delivery account would need to be taken of areas where diminishing returns on technology arise. A segmented approach to technologies might deliver benefits With respect to the commercial market, service could be provided by regional providers. There would be a number of ways in which the commercial market could be serviced That politicians need to think beyond bandwidth That the cost differential between various solutions should be assessed (for instance, connectivity rather than access). What was happening in the rest of Europe and what impact would that have on the UK? That instead of meeting the demand in the easiest areas first, the focus should be on the hardest to reach areas. This would reduce market distortions A request for service under USO should be on a value for money basis which would allow scope for funding the provision in the most optimal way. A range of methods should be considered for the roll out of USO not just copper or fibre but also solutions such as satellite etc That for the development of USO it would be necessary to think outside the infrastructure box There need not be just one business model for USO but a variety of solutions The market needed to be opened for data exchange to be permitted which would then lead to a reduction in costs 7. OSAB members then debated as to whether there should be a USO and whether Ofcom s proposals were in the right direction. The consensus view was that 10Mbit/s was a useful target but that the USO would need to be responsive to changes in technology and use of data. In that respect it would be worthwhile considering a longer term strategy whereby targets are periodically reset. OSAB also noted that the USO should also contain provision for dealing with interference. Item 5: Future mobile services in rural areas 8. OSAB received a presentation on future mobile services in rural areas. The members noted the following points: Goals for the Mobile Data Strategy and 5G o To ensure that sufficient spectrum of the right type is available to meet consumers demands for mobile data

7 o To ensure appropriate spectrum is available to meet the rapid rollout of 5G services (lack of spectrum would be an inhibitor) o To ensure that there is at least one harmonised global band for 5G o To ensure that the UK is a leader in driving spectrum for 5G, if not technology leadership o To ensure that equivalent 5G services are available to all users, and that rural areas do not receive a second class service Assumptions for 5G o The primary focus driving spectrum demand is high speed, low latency services (fat pipes) o Low frequency spectrum for 5G is likely to be from existing bands (such as refarming 800/900/1800 MHz) o The use of new sub-6ghz bands should not be delayed until 5G was available Approach for 5G o Define the 5G services that are key for consumers o If very specific conditions were required to deliver the service in rural areas consider a fundament rethink of the mobile spectrum programme o If they did not what did Ofcom need to think about to deliver services in rural areas o A representative set of 5G use cases and how/whether they could be delivered to rural areas. Services included High speed internet browsing at home/office Video conferencing in every room Smart office Nomadic online gambling Smart buildings/homes (including sensors/ home surveillance) Cloud services along transport routes High quality cloud services everywhere Medium speed internet browsing whilst out and about o The use cases were assessed against benefit and delivery risk in rural areas Consumer benefit what applications are likely to be very niche, and which ones are more likely to have widespread consumer benefit Delivery risk how likely were these applications to be delivered in rural areas. Delivery risk considered by looking at the cost per km² to deliver the service Networks to deliver such services will be characterised by Conventional macro, evolving technologies to improve very wide area coverage. OPEX (site rental) for macro cells was likely to be the main issue High capacity small cells to cover localised

8 use cases, backhaul (and spectrum to allow the aggregation of small cells) was likely to be the main issue Very dense coverage of small cells in urban areas only o Spectrum can help greater mobile reach by providing backhaul and aggregation for small cells o Improved backhaul/aggregation spectrum and technology will present an opportunities for rural areas Wireless fixed links offer an alternative method of backhaul, especially in rural areas where is limited fibre. Microwave links at 11 GHz and above are too big and expensive for small cells. Frequencies below 11 GHz may be especially useful in rural areas, with hop lengths of up to 50 km. cost will fall further as new integrated solutions are developed for small, cheap cells Massive MIMO technologies, which are being developed for 5G, could also be applied to backhaul, increasing range and non-line-of-sight capability (Frequencies at 3-6 GHz may be particularly useful due to the combination of antenna size and propagation characteristics) However rural coverage is relatively low on operators priorities and delivery may be by niche operators/service providers or self-provision by local communities. This may have an implication on how Ofcom awards/authorises spectrum o Actions to enhance availability of wireless backhaul solutions for future 5G services in rural areas include Supporting anything from DCR/BCMR that promotes fibre rollout for backhaul through increased access to BT s ducts and poles Ensuring that mobile spectrum in frequencies attractive for rural backhaul (such as 3-6 GHz) are authorised in a flexible way that allows use for backhaul as well, especially with geographic sharing (eg bands may be needed for access in urban areas, backhaul in rural) Promote the above as a European approach to drive economies of scale Consider the role of alternative delivery platforms (eg satellites/haps for access and backhaul) Consider a campaign to promote future availability of low cost cells with wireless backhaul for rural areas Smaller cells can provide backhaul in hard to reach areas using MIMO in the backhaul Promote technologies that increase reach, not just data rate (eg MIMO can be used for either)

9 Ofcom s role would be not only to make spectrum available but also to influence MNOs, manufacturers, possibly government (re investment on rural broadband) and some rural consumers (eg making information available to users who want to find a solution to a not spot) 9. The members discussed the presentation and offered the following comments: The business model model would be determined by who held the spectrum and who was the operator. It was unlikely that any traditional operator would want to take on digitization of verticals. That 5G could not be separated from a wired environment and that both wired and wireless needed to be addressed That Ofcom might be too conservative in its assessment. Work was on-going in Korea, Japan and the US to make use of the millimetre wavelength and in time new technology will be developed to make use of this resource. That there was a move to a virtualization environment with service providers being wholly virtual. The short range problems for 4G would be solved and there would be a move to a multi global use case where there could be non-human (eg machine) access for use in services such as health That in order to exploit the microwave length there would be a need to have a cell every 50 metres or so and the number required would be in the order of hundreds of thousands That the cost of the new network would not be an issue so long as there was value in the coverage. Coverage itself may improve if there was a change of use. There would also be a need to change the infrastructure in order to achieve a seamless user experience. There was a need to cater for the development of alternative technologies possibly to make innovation driven from the community level That the focus should be at the global level in order to optimise the use of spectrum and eliminate waste 10. OSAB also noted that Ofcom had recently published a consultation on the use of spectrum. The Chairman closed the discussion by noting that this was an issue that OSAB would be interested in seeing again on its agenda. Item 6: Making digital communications work for everyone 11. OSAB received an update on Ofcom s Strategic Review of Digital Communications(DCR). The members noted the following points: That a consultation on Ofcom s strategy for DCR would take place

10 before the summer. The strategy would cover six areas: o Better coverage helping the Government to deliver universal fast broadband and boosting mobile. It was anticipated that coverage would mostly be delivered by the private sector but with public provision o Quality of service a step-change in quality across the industry and automatic compensation o Network choice major shift to promote new fibre networks, through a mixed strategy allowing operators access to BT s network assets o Consumer empowerment better information, service league tables and easier switching o Openreach reform a new governance structure and greater independence from BT to allow for independent strategic decision making o Careful deregulation existing burden of regulation where no longer required or competition is strong. Ofcom would need to decide whether to regulate in favour of the technical savvy or to protect the interests of the disadvantaged 12. The members discussed the presentation and offered the following comments: That there was a grey area between policy and implementation. It was possible that it could result in a longer term Government strategy which would go beyond the lifespan of a Parliament (OSAB were advised that DCMS were fully aware of what would be in the consultation with the possible outcome of a standalone strategy where Ofcom sets expectations for the next 10 years) That the Government would want Ofcom to consider something less focussed on digital communications as it would be looking at the same subject over the next decade That the Government should allow Ofcom more scope to cover areas that it would like to see addressed That the removal of handcuffs from BT s network architecture will maximise the use of assets and encourage innovation That Ofcom needed to determine the longer term destination and work out how to engine of change going without hampering investment (OSAB noted that communication network could not be treated as though it was a utility and that competition would depend upon population density) That if spectrum is not being used efficiently then it raises the question as to whether there is a need for change 13. The Chairman brought the discussion to a close by noting that this would be a subject that would feature on a future agenda. Item 7: Any other business

11 14. A member suggested that a future agenda topic might be discussion on the competition aspects of the Three merger. JJS/MD Chairman