- Hi Tare, please introduce yourself.
I serve as Senior Legal & Regulatory Counsel with SES, a global satellite operator headquartered in Luxembourg. I am based in Singapore and responsible for SES’s government policy and regulatory activities in the Asia-Pacific region.
- What are some challenges you foresee with the upcoming rapid roll out of 5G technologies in the satellite industry?
We see the satellite industry playing an important role in the broader 5G ecosystem. For example, in many places, especially in the Asia-Pacific region, 3G and 4G service is only available because of cost-effective satellite backhaul and trunking.We are getting ready to do the same in the 5G world by investing in High Throughput Satellites (both geostationary and non-geostationary), new ground equipment, and the technical standardization required for ensure seamless integration.
The challenge, we see, is in ensuring we have the spectrum to support the anticipated bandwidth requirements of 5G networks. For example, WRC-19 will consider more than 33 GHz of spectrum for 5G, yet 5G proponents are seeking even more spectrum outside the ITU process in important satellite bands such as the 27.5 – 29.5 GHz (i.e., 28 GHz band) on a country-by-country basis. This is an important band for the satellite industry that has been incorporated into High Throughput Satellite (HTS) systems. In the APAC region, HTS systems already using the band include IPStar, NBN’s SkyMuster satellites, Inmarsat’s Global Xpress satellites, APStar 6, SES-12, and the twenty-satellite O3b constellation. More are expected to be launched later this year by ISRO and Kacific. Many next-generation HTS systems will also use this band (and more!) to meet the growing need for broadband connectivity everywhere.
Fortunately, only a small number of countries (Korea, Japan, Singapore and the U.S.) are proposing to use the 28 GHz for 5G. Most others (the EU-28, China and Australia) recognize that the 28 GHz is already being used intensively by satellite industry and that 5G can be accommodated in other bands. As noted above, WRC-19 will be considering over 33 GHz of spectrum for 5G, and we believe that more than enough can be found among those candidate bands to satisfy 5G demand.
- Can you tell us more about the C-Band Alliance?
In 2017, the U.S. initiated a proceeding to consider using the standard C-band frequencies (3.7-4.2 GHz) for 5G (the so-called “mid-band” proceeding). The pressure in the U.S. to find “mid-band” spectrum for 5G was particularly intense because, unlike most of the rest of the world, the U.S. could not use the 3.4-3.6 GHz band for 5G due to the presence of U.S. government radars in that band.
In response, Intelsat and Intel (later joined by SES and other C-band satellite operators) proposed to clear a portion of the C-band for 5G (about 200 MHz) using secondary market transactions in order topreserve the rest of the standard C-band for satellite use. This would “make up” for the lack of access to 3.4-3.6 GHz in the U.S.The C-Band Alliance (CBA) was formed by the major satellite operators in the U.S. to protect the existing C-band service environment and facilitatethe speedy roll-out of 5G in theU.S., should the U.S. accept the satellite industry’s proposal.
In the Asia-Pacific region, most countries decided at WRC-15 that the use of the C-band frequencies for satellite was paramount and declined to a region-wide identification of any part of the C-band for mobile services. Some 11 countries did, however, indicate that they wished to use some of the C-band (below 3.6 GHz) for mobile services, and those countries were included in a footnote to the ITU Radio Regulations that protects the satellite services in neighbouring countries. Unlike the U.S., those countries have access to “mid-band” spectrum and need not follow the U.S. approach.
- How will 5G’s one-millisecond latency impact satellite applications?
5G applications will have very different latency requirements. In fact, many5G applications will not have the extreme latency requirements, such as Internet of Things, that 5G technologies will enable. In fact, GEO latency is acceptable for many 5G applications, and new MEO and LEO networks will be able to support more latency-sensitive applications.
In addition, as GSMA acknowledged in 2014it will be a major challenge for 5G mobile networks to achieve sub-1ms latency. Interestingly, GSMA’s 2017 5G literature refers to a sub-10ms latency in favour of a sub-1ms latency standard. Regardless, for 5G networks to meet even this revised latency requirementfor the 5G applications that may need it, commonly accessed content must be stored on servers at the edge of the network so that there is no delay in fetching the content. Satellites can help get this content to the edge by efficiently multi-castingcommon content to multiple 5G edge servers across a wide area – not unlike what satellite does for DTH and broadcasting today.
- Considering that 5G requires more base stations to cover the same geographical area as 4G – what does this mean for the satellite industry?
As noted above, the opportunities are potentially immense, especially in a region such as Asia-Pacific where there are many areas that only have mobile coverage today because of cost-effective satellite backhaul and trunking. Here, satellites will play a key role in creating a more inclusive digital society by extending the reach of terrestrial 5G networks to those that would otherwise not be reached. Satellites are very well placed to help achieve the objective of providing affordable, available, and sustainable telecommunications/ICT and broadband connectivity to citizens everywhere, either by supporting terrestrial 5G networks or by directly providing broadband.
- “There is more than enough spectrum for terrestrial 5G in other bands without having to undermine critical satellite connections at 28GHz today” – what is your take on this statement?
This is an important point regarding the 28 GHz that is often missed. The ITU is not considering the 28 GHz band for 5G, but is instead considering more than 33 GHz worth of other millimetre wave bands for terrestrial 5G. It is highly likely that the ITU will identify more than enough millimetre wave spectrum during WRC-19 to satisfy any realistic 5G requirements without needing to look at the 28 GHz band and disrupting the extensive satellite investments and services in the band. There is simply no need or reason to sacrifice satellite services in 28 GHz for 5G when there is some much other spectrum likely to be identified for 5G. .
- (In context of spectrum preservation), what are your thoughts on spectrum sharing?
Spectrum sharing can sometimes a valid policy outcome if anticipated use of the band by any one service will not effectively preclude all other uses and the propagation and technical characteristics of the services allow it. But where at least one of the services anticipates ubiquitous deployment, sharing could result in one of the services being extremely constrained or, worse, every service being constrained so that none can reach their full potential. So, we would encourage countries to take a holistic, balanced approach to spectrum planning that ensures that each service has enough “exclusive” spectrum to reach their potential, and to ensure that, where sharing, is implemented that the conditions for such sharing are fair, equitable and enforceable. Countries should certainly avoid disrupting valuable investments and services in a given spectrum band when there is ample other spectrum to support other services.
- Finally, what are you reading at the moment?
“The Intelligence Trap” by David Robson, which provides a fascinating perspective on intelligence, with lessons on judgment and decision making.