Hong Kong s Role in the Asian International Broadcasting Industry

Hong Kong November 27, 2017 Mr. Tony Lavender Plum Consulting 10 Fitzroy Square London, W1T 5HP (By email) Dear Mr. Lavender, Thank you for reaching out to CASBAA; I m pleased to be able to provide you with this input to the Plum Consulting Consultancy Study on Future Supply of and Demand for Telecommunications Infrastructure Capacity in Hong Kong. Headquartered in Hong Kong, CASBAA is an Asian industry association with members and activities in 17 Asia-Pacific markets. The Association is dedicated to the promotion of multi-channel television via cable, satellite, broadband and wireless video networks across the Asia-Pacific region and represents about 100 corporations, which in turn serve more than 3 billion people. Member organizations with significant operations in Hong Kong include Asiasat, APT Satellite, Asia Broadcast Satellite, Celestial Tiger Entertainment, Fox Network Group, Sony Pictures Television International, Turner International Asia Pacific, TV5MONDE, TVB, 21st Century Fox, and Time Warner. In past decades, Hong Kong has been a regional hub for the international broadcasting sector. It has won this role because of the supportive environment it offers for a dynamic industry which is constantly evolving, in both technical and commercial senses. Hong Kong is also a hub of the Asian satellite industry, and these two sectors have grown symbiotically; satellite transmission is the backbone of the Asian content delivery industry, whether by DTH, cable or broadband. The ongoing growth and development of these twin industries brings major economic benefits to the SAR. This is a dynamic, high-tech sector that will continue to make a very important contribution to Hong Kong s economic growth for the foreseeable future as long as the SAR continues the policies which have favoured the sector s growth. Benign policies related to satellite transmissions, reception, and system operation have been a hugely important ingredient in the SAR s successful regulatory mix. As you will see from the comments in this letter and in the attached questionnaire response, it is important that your assessment of telecommunications industry supply and demand take full cognizance of the vital role played by satellite. Hong Kong s Role in the Asian International Broadcasting Industry While Hong Kong s own broadcasting market is, of course, limited in size, the SAR has played a leading role in regional broadcasting because of its favorable regulatory environment, rule of law and predictable business framework. The government established a licensing regime (including specific provision for non-domestic services) which facilitated establishment and operation of both broadcasting organizations and satellite operators here, and then it stood back and allowed the market to develop in a highly competitive manner without undue regulatory

interference. Spectrum was allocated for these services based on agreed international assignments, and has been sufficient to support sustained growth over the years. As a result, Hong Kong is the home of 3 satellite operating companies having their principal place of business in Hong Kong. These companies own and control 19 operating satellites, whose in-orbit investment value at current costs can be estimated at roughly US$4.7 billion. The satellite companies, teleport operators, and broadcasting organizations (some of which are separately licensed for satellite service) uplink 1 a total of 181 different TV channel feeds to 16 different satellites in the geostationary arc over Hong Kong. These channels supply entertainment, news and sports programming to people throughout the Asia-Pacific region; the relevant audience catchment is enormous, as the region has a population of more than 3 billion people, and more than 500 million households connected to the multichannel television services which are the consumer interface. The vast majority of these services are carried in the frequencies designated as Cband, which have transcontinental reach and relatively high resistance to atmospheric attenuation ( rain fade ). In addition to its supply role as a communications hub and home of the broadcasting industry, Hong Kong s open media regime means there is a very large demand audience domestically which consumes satellite television directly from the satellites (that is, without the intermediation of a cable company or other pay-tv operator). OFCA reports that the SAR s 70 licensed Satellite Master Antenna Television (SMATV) servicers maintain service to about 900,000 TV connections (about 800,000 of which are in residential premises.) These TVs draw on a pool of more than 400 free-to-air satellite channels which are receivable in Hong Kong. The economic and social importance of the international broadcasting industry in Hong Kong is thus very large; when compared to the SAR s small physical size and population, the Hong Kong-based industry s footprint is huge indeed. The connectivity provided by the satellite industry (for broadcasting as well as data uses) is essential to Hong Kong s continuing role as a communications, finance, and trade hub in East Asia. The Role of Satellite and Satellite Spectrum in Future Communications Systems Besides their prominent role in international broadcasting, satellite technologies are also expected to play an important role in the future 5G ecosystem, including: a. By extending terrestrial 5G networks to places they would not otherwise reach, which is essential for a digitally inclusive society b. By efficiently supporting Machine to Machine (M2M) / Internet of Things (IoT) networks through direct connection or backhauling of aggregated M2M/IoT data from multiple locations (e.g. to support 1 As of August 2017

sensor networks and other Smart City applications, or to enable connected cars, planes and ships). c. To help terrestrial 5G networks meet the low latency (<1ms) requirements of some of the new 5G applications through efficient multicasting of commonly accessed content to storage caches at multiple 5G base stations. While most 5G applications (e.g. Internet of Things) will not have low (<1ms) latency requirements, it is projected that a few, still emerging applications might have such requirements (e.g. VR and autonomous driving. According to the GSMA, any service requiring such a low latency will have to be served using content located very close to the customer, possibly at the base of every cell, including the many small cells that are predicted to be fundamental to meeting densification requirements. 2 d. To restore connectivity when existing terrestrial networks have been disabled (e.g. after a natural disaster). Satellites already play comparable roles in today s 2G, 3G and 4G/LTE networks, and are well placed to continue playing such roles for 5G networks, as more High Throughput Satellites (HTS) in both geostationary (GEO) and non geostationary (non GEO) orbits are deployed, and as smaller, more advanced, and lower cost ground antennas are developed. For satellites to play their role, they will need continued, sustainable access to satellite spectrum. Many HTS satellites have already been deployed in multiple bands, including the C-band, Ku-band and the mmwave band spectrum. For example: Intelsat s EPIC, Eutelsat s Eutelsat-172B and APT Satellite s APStar 5C HTS designs use the C- and Ku-band frequencies; Thaicom s IPStar and SES s SES-12 HTS satellites use (or will use) the Ku- and Ka-band frequencies; and Inmarsat s Global Xpress and the Kacific-1 HTS designs use the Ka-band frequencies. Many more HTS systems are being planned, including new nongeostationary constellations, such as O3b (Ka-band), OneWeb (Ku-/Ka-band), and SpaceX (Ku-/Ka-), just to name a few. CASBAA urges governments to avoid the spectrum bands that are already in use or that are likely to be used for HTS (and future VHTS) satellite systems. These include: a. 26 GHz Band (24.75 27.5 GHz). The 24.65 25.25 GHz band is identified by ITU as the Fixed Satellite Service (FSS) uplink band intended to feed the 21.4 22 GHz Broadcasting Satellite Service (BSS) downlink band. The ITU affirmed and expanded this allocation as recently as WRC 12. Without access to this uplink band, the ability to efficiently feed the 21.4 22 GHz band is lost. Satellite operators are just starting to deploy in this band in other parts of the world (e.g. DIRECTV 14 & 15). In the future, it is conceivable that the BSS bands would be used to efficiently multicast content to multiple 5G base 2 See GSMA Intelligence, Understanding 5G: Perspectives on Future Technological Advancements in Mobile, at 12 (Dec. 2014).

stations, as described above. Some HTS systems have also begun to deploy in the 27.027.5 GHz band (e.g. Australia s NBN). Governments should either avoid use of these bands for 5G or ensure shared use of the band by satellites and 5G mobile services. Sharing is currently being studied by ITU R TG 5/1 and the results of those studies could require constraints on terrestrial 5G systems. b. 28 GHz Band (27.5 29.5 GHz). This band is already extensively used for the latest and next generation HTS systems around the world, both GEO and non GEO. Tens of billions of dollars are already invested or planned to be invested in GEO and non GEO satellite systems at 28 GHz. As a result, all regions at the ITU World Radio Conference 2015 (WRC 15) decided not to study this band for 5G/IMT 2020 at the next World Radio Conference (WRC 19 Agenda Item 1.13). c. 37-42.5 and 42.5-52.6 GHz Bands (Q- and V-bands). Multiple satellite operators have planned next-generation Very High Throughput Satellites (VHTS) using parts of this band, including at least six large constellations of non-geo satellites, e.g. Boeing, O3b/SES, OneWeb, SpaceX, Telesat, and Theia). At the very least, governments should consider limiting their focus to the parts of these bands (37-40.5 GHz, 42.5-43.5 GHz, 45.5-47 GHz, 47-47.2 GHz, 47.2-50.2 GHz, and 50.4-52.6 GHz) that are being studied for 5G/IMT- 2020 under WRC-19 Agenda Item 1.13. Sharing studies between future 5G/IMT-2020 and satellite services in the WRC-19 Agenda Item 1.13 bands are underway at the ITU and should be taken into account once concluded. To the extent that future 5G/IMT-2020 mobile and satellite services are to share the same spectrum, then reasonable sharing conditions based on these studies will need to be developed. A number of other mmwave in higher frequency bands will be considered for 5G / IMT-2020 terrestrial mobile services under WRC-19 Agenda Item 1.13, including the 31.8-33.4 GHz (32 GHz), 66-76 GHz (66 GHz) and the 81-86 GHz (81 GHz) bands. There is adequate spectrum in these bands to meet terrestrial 5G requirements without the contention with satellite spectrum that is foreseeable in the Ka-, Q- and V-bands. The 66 GHz and 81 GHz bands, in particular, are considered very good prospects for international harmonization given their limited existing and planned use by other radio services. The 66 and 81 GHz band in the high mmwave bands should yield about 15 GHz of spectrum in contiguous blocks of at least 5 GHz, which could support very wide-band 5G/IMT-2020 carriers. These high mmwave bands should therefore be able to support the development of 5G mobile networks in highdensity indoor and outdoor scenarios, such as stadiums, campuses or shopping malls located in urban and suburban areas. The use of these bands would also benefit from synergies with WiGig currently being deployed at 61 GHz for which chipsets and MIMO antenna systems are already being manufactured. Attached you will find more detailed responses to the questions you posed in your questionnaire. (We have not attempted to answer all questions, as only some are relevant to our industry.) I should also note that this information and these responses are offered on behalf of the international satellite and broadcasting industries; Hong Kong s domestic telecommunications operators may have their own

views to offer and we trust you are engaging with them separately to obtain information from them. Sincerely yours, John Medeiros Chief Policy Officer CASBAA

Responses to the Plum Consulting Questionnaire 2.1 Demand 1. How has demand for local and external telecommunications services changed in the past 10 years in terms of: Retail demand from customers / subscribers (consumer, enterprise1 1 and government). Wholesale demand from service providers (e.g. MVNOs), inter-network infrastructure demand (e.g. for interconnection / peering between networks) and demand for access to external infrastructure including terrestrial, sub-sea and satellite). 2. How do you see demand for local and external telecommunications services changing in the next 5-10 years (2017-2027) in terms of: Nature of use (e.g. voice communications, high-definition video, data transfer, etc) Number of users, connected devices Geographic requirements (e.g. coverage, location specific needs) Bandwidth requirements per user or device (e.g. MB/user/month, uplink vs downlink needs) Daily demand profile (e.g. busy hour traffic, usage) Technology/network requirements (mobile, fixed, Wi-Fi, satellite, submarine cable, overland cable to the Mainland, etc) Quality and latency of service, resilience requirements (e.g. uptime, availability, etc) for local and/or external telecommunications services 3. What are the demand drivers and key market trends which will shape your ICT needs over the next 5-10 years (e.g. 5G, OTT, IoT/M2M, VR, 4K video/tv, smart city and cloud-based technologies, services and applications)? For the Asia Pacific region, video distribution will continue to be the major satellite capacity user over the next decade. Absolute numbers of channels will not increase as rapidly as in the previous decade, and may stabilise overall over the period. However, the shift to higher definition services whether simulcast or not, and whether from SD to HD or from HD to 4K UHD (and in Japan and Korea 8K UHD) should buoy demand throughout the period. VR is not expected to generate significant proportions of demand in the next five years. The subscriber bases of DTH satellite services in China and India will continue to grow, though more slowly than in the previous decade. In India, some consolidation of DTH operators is expected, which may reduce demand for DTH-suitable satellite capacity there. However, growth in the number of TV channels distributed by 1 Note that enterprise customers will include entities with data centres and similar infrastructure, which may require very high capacity connection.

satellite over India is expected to continue, as services are further developed to meet the needs of the country s diverse language speakers. Hong Kong itself no longer has any inbound licensed DTH pay-tv services. Hong Kong consumers receive satellite TV on regional unencrypted services. Other satellite services delivered into and out of Hong Kong include services for: aeronautical and maritime, financial sector, oil and gas, bilateral voice for unspecified markets and video, among others. 5G is unlikely to make a significant impact in Hong Kong or most other markets until at least five years from now. The demand for the different types of 5G services is not established clearly. According to a number of forecasts, the initial demand will be for fixed wireless access i.e. providing fixed ultra-fast broadband services to premises not reached by fibre. In the meantime, 4G consumption and capacity are highly driven by video, and 4G/5G data forecasts indicate that the proportion of data used to convey video will increase to over ¾ of all data in the next five years: e.g. Cisco VNI 78% for APAC in 2021, Ericsson Mobility Report 75% worldwide in 2022, Huawei-Ovum The Evolution of Big Video 75% worldwide in 2021. Moreover, an increasing proportion of the video is expected to be live video, most of which means broadcast (as compared to Skype, WhatsApp video calls etc). Cisco VNI forecasts: Live Internet video will account for 13 percent of Internet video traffic by 2021. Live video will grow 15-fold from 2016 to 2021. 2 In addition to all the live video, a substantial portion of video and other content (e.g. popular mobile operating system updates) is very frequently requested: head content; as compared to the much wider long tail selection of content available. While unicast serves the long tail content well, multicast and broadcast are much better ways to deliver the head content, both directly to reception devices for live consumption (or time-shifted consumption, subject to storage in the reception device), and to edge caches in the network. The same Cisco VNI paper linked above forecasts: Seventy-one percent of all Internet traffic will cross CDNs by 2021 globally, up from 52 percent in 2016. 2 Satellite is an excellent, cost effective way to distribute live and head content over large, sparse, and disconnected areas. Other applications of satellite to 5G are discussed in Sections 2. and 3. of the joint satellite associations response to China s Ministry of Industry and Information Technology s (MIIT) Open Solicitation of Feedback on 5G Communications Systems in the MillimetreWave (mmwave) Bands attached as Annex A. 2 https://www.cisco.com/c/en/us/solutions/collateral/service-provider/visualnetworking-index-vni/vni-hyperconnectivity-wp.pdf

OTT video is often misconstrued to be the killer of broadcast TV, but statistics to date in developed markets indicate this portrayal is gross exaggeration. Legitimate OTT operators such as Netflix are acting more and more similarly to established multichannel distribution platforms. For the first time in years, there is real evidence to show that with the right combination of policy, law, technology, policing and legal process, significant headway can be made against online OTT video piracy, which has in recent years dented demand for legitimate multichannel and pay TV services 3. 4. Are there any aspects of demand which are not being met at present? 5. What impact would the new technologies and services mentioned in Question 3 of this section have on demand for telecommunications infrastructure? 6. What are the factors or barriers that could inhibit demand over the next 5-10 years? 2.2 Supply 1. How has supply of local and external telecommunications services changed in the past 10 years in terms of: Retail supply to customers / subscribers (consumer, enterprise and government). Wholesale service providers (e.g. MVNOs), inter-network infrastructure demand (e.g. for interconnection / peering between networks) and demand for access to external infrastructure including terrestrial, sub-sea and satellite). 2. How do you see supply of telecommunications services changing in the next 5-10 years? What is the technology roadmap for telecommunications services (mobile, fixed, Wi-Fi, satellite, submarine cable, overland cable to the Mainland, etc) over the next 5-10 years? CASBAA will limit this response to satellite, since this area is of most direct interest to its members. Remarkably for its size, Hong Kong has two domiciled and Hong Kong listed satellite operators, APT Satellite and AsiaSat. Hong Kong is a major operational base for a third operator, ABS. The three companies have a total of 19 satellites currently in service, 13 of which serve Hong Kong and surrounding territories. All three operators are in the top 20 satellite operators worldwide in terms of revenues generated. Hong Kong is served by numerous other satellites in the geostationary arc from approximately 45 East, to 185 East, including satellites operated by Intelsat, SES, Eutelsat, SKY Perfect JSAT, Thaicom, China Satcom and Measat. 3 http://www.dailymail.co.uk/sciencetech/article-5070091/sky-boss-claimsfirm-winning-war-against-online-piracy.html

Responses gathered by CASBAA from four operators: APT Satellite, AsiaSat, Intelsat and SES, indicated current transponder capacity covering Hong Kong from these operators alone of: 387 C-band, 418 Ku-band and 2 Ka-band transponders, with transponders in all bands expected to increase in the next few years, especially Kaand Ku-band. Some satellite and related developments and trends include: High throughput satellites (HTS) using multiple concentrated spot beams, with coverage areas of the order of 100 times smaller than regional beams, a high degree of frequency re-use, and in some cases ultra-wideband transponders, these can achieve 20 times greater throughput (e.g. 30 100 Gbit/s) and lower cost per bit than other satellites, enabling cost effective, high capacity data communications in underserved areas, to air, land and maritime mobile, 4G mobile backhaul services, as well as international telecommunications and video distribution. A number of HTS s are serving Hong Kong and the wider region today, operating in C-, Ku- and Kabands, including: IPStar, O3b, Inmarsat GX and Eutelsat 172B. In the next two years HTSs named APStar 5C, SES-12, JCSAT-18 and Horizons-3e will also be launched to serve the region and HK. Low-earth orbit satellites (LEO) at altitudes of 160 2,000 km (c.f. 35,786 km for geostationary) and corresponding orbital periods of 84 127 minutes, these can cover high / low latitudes not practically accessible with geostationary (GEO) satellites, require lower power transponders and cheaper ground equipment due to 1/300 or less times the path loss, enable lower latency communication and are cheaper and faster to produce using mass production techniques and to launch. However, many more are needed OneWeb will have a constellation of 900 satellites the lifetime is much shorter than the 15-20 years now typical for GEO satellites, and the complexity of frequency coordination and operation of LEO satellites is much higher than of GEO platforms. Medium-earth orbit satellites (MEO) or intermediate circular orbit (ICO) satellites orbit at altitudes between LEO and geostationary and orbital periods 127 minutes to almost 24 hours. Used by the first communications satellite, Telstar, and established almost 40 years for GPS; like LEO, this is not a new type of orbit. MEO communications satellites offer a middle-way between GEO and LEO, with much smaller constellations than LEO required for comparable service coverage. For example, the current O3b MEO constellation comprises 12 satellites in equatorial orbit, covering latitudes from 45 North to 45 South initially, and this will be increased to 20 satellites, with some in inclined orbits, extending coverage to polar regions 4. Combining use of LEO and GEO or, as in the case of SES and O3b, MEO and GEO constellations gives operators even greater flexibility to offer better optimised services. 4 http://spacenews.com/ses-takes-control-of-o3b-citing-multiple-synergiesbetween-geo-meo-constellations/

Electric propulsion ABS operates two all-electric satellites the first of their kind in commercial use. All-electric propulsion for orbit raising and station-keeping fuel both reduces launch mass and cost, and extends the usable lifetime of the satellites. Other satellites lifetime is generally determined by the supply of station-keeping fuel, fixed at launch. Wideband transponders compared to using multiple (say) 27, 36, 54 MHz transponders, using a smaller number of (say) 500 MHz transponders reduces satellite mass and launch cost. Used with appropriate wideband modulation, wideband transponders can also make more efficient use of the spectrum as a far lower proportion of spectrum is lost to guard bands between the adjacent transponder bands. Previously, use of wideband transponders necessitated multicarrier operation, which itself both reduced the bandwidth efficiency, due to guardbands required between carriers, and reduced the power efficiency, due to the need for transponder backoff to operate the travelling wave tube amplifiers linearly. Use of advanced modulation (see below) and wideband tuner / demodulators in ground based reception equipment enables very high efficient bandwidth and power utilisations to be achieved with wideband transponders. Advanced, high efficiency modulation DVB-S2X DVB-S, standardised in 1994 is still in use today for DTH SD video distribution in many markets. Since 2005 many operators adopted DVB-S2 along with MPEG-4 AVC for delivery of newly launched HD video services, requiring new set top boxes. DVB-S2 enables up to 36% more bit rate capacity in a given DTH transponder than DVB-S, or up to 280% more for non- DTH applications. Notably, two leading DTH operators in India that had launched with DVB-S, upgraded all their transmissions to DVB-S2, replacing many millions of subscriber set top boxes in the process. DVB-S2X, standardised in 2014, is intended as a more general-purpose modulation solution, extending the usability from very power constrained links to very bandwidth constrained links, supporting more diverse applications including air mobile, very small antennas, cellular backhaul and cruise ship networks, plus primary video distribution to terrestrial video networks and DTH. DVB-S2X enables new capabilities including channel bonding the ability to bond multiple transponders to carry very high data rate payloads multiplexed across several transponders and wideband transponder support. DVB-S2X allows up to 30% more capacity in standard non-dth transponders than DVB-S2 5. Higher frequency spectrum bands apart from C-band, which was already used for terrestrial microwave in certain territories before the first satellites were launched, satellite manufacturers and operators have led the way in commercialising use of higher frequency bands, such as now well-established Ku- and Ka- bands. The next decade is likely to continue this trend. For example, at least six satellite manufacturers and operators have filed with the US FCC for LEO and MEO satellite 5 http://www.agfranz.com/product-docs/teamcast/satmagazine-2016oct- TeamCast-Annex-M.pdf

constellations using V-band frequencies 6 V-band is the next higher frequency band above Ka-band, spanning approximately 40 to 75 GHz. Note that each of the six companies is filing for specific, much narrower bands of frequencies toward the lower end of the V-band range. 3. What are the market trends that could influence supply (e.g. competitors, technology such as 5G mobile technology, new services/applications such as OTT, IoT/M2M, VR, 4K video/tv, smart city, cloud-based services/applications, new entrants)? 4. What are your views on the phasing out of legacy ICT systems, networks and equipment (e.g. 2G/3G, copper, analogue TV) in the next 5-10 years? What are the factors and issues that will affect the phasing out of legacy such systems, networks and equipment? Peculiarities of the local market may mean that 3G is phased out around the same time as, or even before 2G 7. Many existing IoT and M2M services are well served by the cheap, highly reliable, ubiquitous 2G services, but at some stage it will become uneconomic for service providers to maintain 2G services. Digital terrestrial rollout commenced at the end of 2007 after the government s policy decision on DTT implementation framework in mid 2004. Analogue switch off in Hong Kong has twice been postponed: from end 2012 to end 2015, then to end 2020, although even this is subject to a review to take place in 2017/18. Although DTT coverage is now 99% or better 8, OFCA last updated its figure for the DTT take-up rate, 85%, almost two years ago 9. It is clear that the government does not see freeing this 108 MHz spectrum (698-806 MHz) for 4G / 5G wireless as a priority, as it is not even tracking the DTT take-up rate in a timely manner, let alone trying to encourage DTT take-up. This is galling in light of the government s insistence that it needs to recover other frequencies in active use by the satellite industry and specifically for satellite operators and their customers using capacity and services in the extended C-band, 3.4 3.7 GHz, which OFCA seems to be about to put on less than three years notice to vacate in Hong Kong. Please refer to the answer to question 7. below. 5. What role do you see the new technologies and services mentioned in Question 3 of Section 3.1 above playing in terms of meeting the needs of different types of users? What is your view on the likely take-up rate of each of these technologies and services? Please provide reasons. 6 http://spacenews.com/fcc-gets-five-new-applications-for-non-geostationarysatellite-constellations/ 7 https://www.mobileworldlive.com/blog/blog-who-will-be-the-last-to-turn-off- 2g-or-3g/ 8 http://www.digitaltv.gov.hk/general/ 9 http://www.ofca.gov.hk/en/media_focus/data_statistics/key_stat/

6. In terms of wireless connectivity, what roles do you see Wi-Fi and other licenceexempt technologies (e.g. IoT, RFID, Bluetooth) playing in delivering telecommunications services over the next 5-10 years? 7. What are your views on the radio spectrum roadmap for Hong Kong in the next 5-10 years, taking into account the Communications Authority s work plan announced in March 20172 10? Do you anticipate new models of spectrum supply (e.g. dynamic sharing) over the next 5-10 years? How would this affect supply of telecommunications services? CASBAA and its satellite members oppose the re-allocation of the extended C-band 3.4 3.7GHz to mobile services, because several operators, including Hong Kong based operators, provide capacity in that band that is in use both in and outside Hong Kong today. The effect of vacating that band in Hong Kong would in reality be to deprive Hong Kong residents of the ability to receive certain content, because broadcasters and channel distributors are unlikely to want to cause disruption to all their other receivers of the same content outside Hong Kong for the sake of their Hong Kong audience. The CA s workplan includes a push for hasty clearance of these bands by 2020, i.e. less than three years notice. This, even though the band is being efficiently used for services including multi-national TV distribution services into Hong Kong SMATV systems, and even though relevant satellite licenses permitting use of spectrum in that band were issued to Hong Kong based satellite operators less than three years before the apparent change in direction. This is particularly galling when compared with the exceedingly slow vacation of the 700MHz band by long-legacy, obsolete analogue TV, as explained in the answer to question 4. above. OFCA issued a subsequent Consultation on Proposed Change in the Allocation of the 3.4-3.7 GHz Band from Fixed Satellite Service to Mobile Service in July. CASBAA, its satellite members and related satellite associations opposed the re-allocation. CASBAA s response is attached as Annex B. It refers to the joint satellite associations response to China s Ministry of Industry and Information Technology s (MIIT) recent Open Solicitation of Feedback on 5G Communications Systems in the MillimetreWave (mmwave) Bands attached as Annex A. OFCA has not yet published consultations on the 26 GHz band (24.25-27.5 GHz) or the adjacent 28 GHz band (27.5-28.35 GHz). However, the position of CASBAA, its satellite member companies and other satellite associations is to oppose their 10 The Communications Authority announced its workplan to make available additional spectrum for public mobile services in March 2017 (see http://www.comsauth.hk/en/media_focus/press_releases/index_id_1423.html) and published a consultation on the proposed reallocation of the 3.4-3.7 GHz band from fixed satellite service to mobile service in July 2017 (see http://www.coms-auth.hk/filemanager/en/content_711/cp20170727_e.pdf).

allocation to 5G mobile services, for the reasons clearly set out in Annex A Sections 4. a. and 4. b. respectively. 8. Would supply be able to meet the requirements of Hong Kong s Smart City Vision and critical infrastructures of the business/commerce/industrial sectors (including telecommunications service providers), public bodies or organisations and the Government? Please state your reasons. 2.3 Other factors 1. Are there other factors which may influence demand and supply (e.g. new technologies/services/ applications other than those mentioned in Question 3 of Section 2.1 above, telecommunications market competition, policy or regulatory developments, investment constraint, wider economic climate, digital skills)? 2. What impact would Hong Kong s Smart City Vision have on demand and supply of telecoms infrastructure? What are the main challenges in achieving the desired outcomes? At this early stage of the Vision / Blueprint, it is difficult to forecast its specific impact on satellite capacity demand and supply. However, satellite operators have roles to play in Smart City infrastructure and services. Satellites to Provide Vital Link to Connected Cars, attached as Annex C, explains as an example one role that is highly applicable to Hong Kong s Smart City Vision / Blueprint. 3. What are your views on the Hong Kong telecoms market in relation to other developed markets internationally? Are there any aspects which can be improved? The spectrum policy framework developed in 2007 11 has not been substantially reviewed or changed since then, nor have certain developments envisaged back then e.g. spectrum trading and spectrum liberalisation (for use of different technologies, subject to non-interference criteria being met) been implemented. Section 1.2 of the spectrum policy framework states: This document is a living document and will be reviewed, revised, modified and updated from time to time by the Government as it thinks fit but it seems it has not thought fit in the past decade. The policy, implementation and systems need to be updated and the spectrum regulation and allocation considerations and priorities need to be more transparent. The policy principle that there should be no legitimate expectation for renewal of spectrum assignments, coupled with relatively short by developed territory standards license periods appear to be the source of much of the uncertainty, 11 http://www.cedb.gov.hk/ccib/eng/legco/pdf/spectrum.pdf

which hinders investment, holding back Hong Kong s broadcasting and telecommunications as compared to other territories. At very least, spectrum allocated should remain allocated for the duration of relevant licenses, except in the most exceptional circumstances e.g. force majeure, prolonged failure to provide a minimum level of rollout, service or utilisation without good cause, repeated marketing fraud, major financial irregularities. The impression given at present is that certain narratives (e.g. 5G, Smart City ) currently attract or will attract more favourable regulation and allocation than existing revenue- and tax-generating services provided to or benefitting Hong Kong residents. Moreover, it appears that the government is willing to sacrifice many of the latters interests today for the promise of greater things in future, based on optimistic demand forecasts attempting to justify significant spectrum grabs. Due to its multi-territorial nature especially when seen from within a geographically small territory like Hong Kong satellite spectrum does not lend itself to being allocated by auction, as mobile spectrum has been over the past 16 years. Moreover, with retrofitting of geostationary satellites not practicable, and service lifetimes of 15-20 years now typical, licenses granted should be honoured for at least 20 years from launch.