Digital Transformation Monitor

Low-Earth Orbitsatellites:Spectrum access

July 2017

Internal Market,Industry,Entrepreneurshipand SMEs

2

Low-Earth Orbit satellites:

Spectrum access

Traditionally, geo stationary satellites located at more than 36,000 km above the earth have been used to providesatellite communication to a wide area. An alternative approach however, once unsuccessfully tried in the past is nowsurfacing again. The idea is to use considerably massive constellation of satellites at lower altitudes to both densifythe network and reduce latency. Although technically feasible, those numerous projects come with challenges andaccess to spectrum is one of them. At stake, the possibility to provide enough capacity at the right price points to servemarkets that remain up to now unserved.

The cost of such solution howeverremains important in the absolute andsolutions are regularly looked aftertomake this solution more and moreaffordable

Already tried in the past but more

prone to succeed today

In the 1990’s, several players came withthe idea of launching a multitude ofsatellites in low Earth orbit (LEO) withthe aim of providing communicationservices at low cost.

Instead of using only a few highorbit satellites, especially geostationaryorbits where satellites have fixedpositions relative to the Earth (at anelevation of around 36,000 km), the ideawas to place smaller and lighter satellitesorbiting at lower altitude (in the range of180 km to 2,000 km) and thus drasticallyreduce the cost of the solution because ofsatellite less expensive to manufactureand launch. This however came withchallenges, both technical and financial.

Why are Low EarthOrbit gamechangers?Because of their high position above theearth, satellites are very effective inproviding wide coverage. Indeed whenone cellular antenna on earth covers onlya few square km, one satellite, dependingon its elevation covers thousands ofsquare km.

This explains why, when terrestrialinfrastructures have not been deployed,satellite is usually seen as a bettereconomic and practical solution to bringcoverage to areas that would otherwisebe left completely unserved.

1Above all, mobile terrestrialcommunication systems were about tobe launched all over the world and all ofa sudden what was supposed to be arevolutionary service was turned into amore expensive and technically lessinteresting offer than the then explodingGSM. In the end, few low Earth orbitconstellations were launched and at thebeginning of 2000, most LEO projectshad been dropped or had gone bankrupt,only to be purchased by investors at afraction of the initial cost.

After that, it became pretty clear in theindustry that the geostationary satellite,with its wide-coverage capacities and itsmore recent high throughput systemcapabilities were the solution toproviding communication services in aworld with demand increasing by theday for broadband speeds anytime,anywhere.

Fundamentally however the veryconcept of LEO communication satelliteswas not dead. While the economicinterest of providing voice andnarrowband coverage from a lowelevation satellite position was notdemonstrated, the idea of using LEOposition to provide broadband accesssurfaced few years ago with severalannouncements from yet unknownplayers at that time. Technologicalprogress had been made and the marketprospect also had changed. If onesatellite with a very broad coveragemakes sense for narrowband usage, is itpowerful enough to provide broadbandcapacity for a market that is still largelynot served?

Figure 1: Presentation of LEO, MEO and GEO orbits

Source: Communication satellites by William Stalling (2001) (1)

© 3Dsculptor/Shutterstock.com

3

Figure 2: Main broadband-focused LEO constellations

Source: IDATE

Similarly, with a densified network,capacity is much more increased becauseof more important frequency reusecapacity. By adding more satellites in theconstellations, capacity and as a resultthroughput per user can be easily addedat a marginal cost.

Who makes what? Which players,

which markets?

Several new entrants not linked to

traditional satellite players

Among the more than 15 LEO projectsthat have surfaced in the last 3 years, anumber of new players is vying to launchnew constellation projects focusing on amarket where high capacity and highthroughputs are required.

The three most publicized projects inthis regard are OneWeb, LeoSat andSpaceX initiatives. The latter hasreceived much press because of its CEOElon Musk, also funder of Tesla, whoenvisions sending people to Mars.SpaceX is also known for thedevelopment of a launcher, whose 1ststage can be reused (3).

This capability is aimed at furtherreducing the cost of launching satellite tospace. The fact that these projects havebeen the most talked about does nothowever necessarily mean that they are

Low-Earth Orbit satellites: Spectrum access

the most likely to succeed, except maybefor OneWeb, whose status is already welladvanced with 1.7 billion USD of fundingalready secured.

Globally, the full completion of all theLEO projects would result,approximately, in more than 17,000satellites in various LEOs providing atotal aggregated capacity of morethan150 Tbps. The main question that israised here is whether there are enoughplaces on the market for all thoseplayers.Most probably, only a fewplayers will succeed in launching theirconstellation.

Backhauling, consumer and enterprisebroadband as main markets

Given their characteristic, LEO areparticularly relevant for use cases wherelarge capacity is required. Notsurprisingly, most ambitious projects intheir size focus on broadbandapplications, something that can beserved with different approaches:

Through mobile backhauling: in areaswhere terrestrial backhauling is toocostly to deploy.

Fixed Broadband access for B2C inremote/unserved/underserved areasincluding in emerging markets.

Broadband for high mobilityapplications (trains, planes, boats).

Dedicated governmental, scientific andenterprise connectivity.

M2M and imaging targeted by otherplayers

While broadband focused projects havegotten much of the highlight, M2M, assettracking and imaging are also targetedby other kind of players.

The only three LEO projects of the 2000sto have survived after bankruptcy(Iridium, Orbcomm and Globalstar) havealready launched a second generation ofsatellite to replace the first one. They,still focus on M2M and relativelynarrowband applications (voice,messaging). Those projects thus requireless satellite. Iridium NEXT for instancewill have 66 operational satellites.

As for imaging, this market has beeninvested by new players, often start-upsleveraging very low cost micro satellites.Planet is one of them. It aims atphotographing the earth every day andthus take quasi real time imagery. Thoseupdates will be critical for cartographyand will pave the way for new innovativeapplications based on low cost earthobservation.

LEO: Densifying the network to bringincreased capacity and reducedlatency

Reduced coverage as compared to GEO…

The elevation that characterizes LEO hasa direct impact on coverage. At anelevation of 36,000 km, coverage from asatellite is 2.7 wider than it is at 1000kmabove the surface of the Earth. Whenthree satellites are required to cover theEarth on GEO, at least 15 are required tocover the whole Earth on LEO. Sincesatellites at this elevation are moving,many more are required to provideconstant coverage.

… But reduced latency and increasedcapacity

But the lower elevation of LEO satelliteconstellation also has benefits. Indeed,since satellites are closer to the Earththan GEO satellites, the trip timebetween the ground and the satellite andthe satellite and the ground issignificantly reduced and so is thelatency, which is one of the caveats ofGEO satellites in general. Latency iscritical for real time applications, ofwhich VoIP notably.

280 vs 20 ms (2)

Typical latency of GE0 vs LEO:

17,000 satellites for morethan 150 Tbps

4

Is 5G an opportunity or a threat forthe satellite industry?

The biggest competition however maynot come from the satellite industry butfrom the further expansion of terrestrialcommunication itself through 5G. If LEOprojects can help decrease the cost ofsatellite connectivity significantly, it stillcannot compete with terrestrial networkwhen fixed infrastructures are alreadythere (backbone and backhaul). This is athreat for the viability of LEO projects.

Using satellite and especially LEO as abackhauling technology will be a way tocomplement rather than compete withsuch movement should operator reallydecide to bridge the connectivity gap.Because 5G will use frequency bandssimilar to satellites (mmWaves), satellitecould also be more tightly integrated into5G. A 5G device would for instanceseamlessly switch to any satellite

connectivity available once connectionwith terrestrial infrastructure lost. Thiswould enable the persistent coveragegoal of 5G. Asset tracking is a goodexample of service that would benefitfrom this capability. A plane that wouldbe connected to an Air To Ground 5Gnetworks would similarly benefit fromthis capability when crossing the oceans.

This kind of complementarity iscurrently limited. Solutions exist but areoften proprietary and costly, hence therequirement for a, cooperation betweenterrestrial and space sector. This needsto materialize in the standardizationspace, especially within 3GPP wherecellular technologies are defined but alsoon the regulatory field with the spectrumallocations decided on the internationaland local level.

For cooperation to be possible orcompetition to be fair, each player musthave a fair access to spectrum dependingon its need and the value it bring to themarket / society.

Which impact of LEO constellationprojects on the satellite industry?

Because most ambitious LEO projects infigures are carried out by new players(i.e. not traditional GEO operators), LEOis often seen as a potential disruption inthe industry together with thetechnologies it involves. With theircapability to drastically reduce the costto deliver 1 Gbps, LEO could indeedfurther increase a price competition thatis already installed in some markets andis causing financial difficulties even tobiggest satellite operators on the market(Intelsat, Eutelsat, SES…)

If we look at estimated CapEx to deliver aGbps, we indeed see a sizeable differencebetween OneWeb for instance and atraditional GEO satellite operator. WhenOneWeb would invest 260,000 USD toprovide 1 Gbps, it takes between 2 and 9million USD for a traditional GEOsatellite operator.

The race toward cheaper satelliteconnectivity is however not over. GEOconnectivity price has already welldecreased in the last years with thedevelopment of HTS satellites andseveral innovations will enable GEOcarriers to reduce cost even more. A GEOsatellite such as ViaSat-3 due to belaunched in 2019 should thus provide a1 Tbps capacity. With 3 satellitesrequired to cover the earth, the cost toproduce a 1 Gbps capacity could be wellbelow the 1million USD price point.

Low-Earth Orbit satellites: Spectrum access

Figure 5: Market potential compared, by technology in low-density areas

Source: IDATE

As another comparison point, Eutelsatexpect to lower its CapEx required toprovide 1 Gbps to this same 1 millionUSD threshold in the mid term (4).

Complementarities between LEO andGEO exist

However, confrontation between orbitsholds true only if LEO, MEO and GEOplayers remain distinct. It does not nowseem to be the case as almost all GEOplayers are currently involved directly(building) or indirectly (partnership) inLEO / MEO activities. Of the top 5 fixedsatellite operator, only Eutelsat hascurrently no implication in such projects.

In a competitive environment, LEO andMEO are increasingly seen ascomplementary to GEO and as a way todifferentiate from the competition. IfLEO seems more adequate to providehigh capacity and low latency for mostdemanding applications, GEO remainsthe most adapted for broadcast purposein a context where video still account forthe most important part of revenues inthe industry.

2

How will LEOconstellations farewith competition

Figure 4: OneWeb vs ViaSat 3

Source: IDATE

Figure 3: Evolution of GEO capacity in Gbps

Source: IDATE

5

Low-Earth Orbit satellites: Spectrum access

The more spectrum, the morecapacity and bandwidth

Frequency bands are of course of criticalimportance for any wirelesscommunication system whether on earthor in the space. This is all the more trueas more and more spectrum are requiredtoday to increase throughput andcapacity. In the past only limited amountof spectrum was required to providevoice or messaging services. With thedevelopment of new data intensiveusages such as video consumption,access to the cloud, holding a limitedamount of spectrum is no longer anoption and most players are fighting toget this access.

Not all frequency bands are equal…

It is important to understand that not allfrequency bands have equalcharacteristics. Therefore frequencybands have different values to players.The position of the band in the spectrumis important but the width of thefrequency band is important too. Beloware basic principles that need to beunderstood:

Figure 6: Frequency bands and main areas of use

Source: IDATE

Lower part of the spectrum favourspropagations. It is better for instancefor indoor coverage or rural areas forterrestrial players. For satellites it isalso better in places where rain fallsare important, even though mitigationtechniques have been developed

In turn, higher frequencies are moredifficult to exploit but as a result havemore bandwidth to use. Satelliteplayers have been historically capableof using higher frequency becausemore power can be used fortransmission and thus compensate forsignal degradation,

Higher frequency bands come withsmaller beams. Wider beam are betterfor broadcast usage while smallerbeams are better for capacity.

Ka band is the preferred band forcapacity but its access could prove tobe difficult in the future

To facilitate the identification offrequency bands, letters have beenassigned to them. The main bands usedby satellite players are presented in thefollowing figure, ordered according totheir position in the spectrum.

This also corresponds historically to thetime when those bands started beingmastered and used. Ku band has beenused for many years, while Ka bandusage, although growing rapidly is muchmore recent.

Because they provide enough capacityand can be used with smaller antennasbut also because satellite operators nowhave enough experience with them, Kuand above all Ka band are the preferredbands for LEO constellations projects.Ka band is already the preferred bandfor GEO High Throughput Systems (HTS)with the notable exception of Intelsatwho prefers the Ku band because it isless subject to interference created byrain falls.

This means that Ka band is currentlymuch looked after and that using thisband will be increasingly complex in thefuture in order to limit interference. Theprospect of 5G networks to use part ofthis frequency band is a concern forsatellite players. In the US, but also inSouth Korea, it has already been decidedthat the 28 GHz band, located in the Kaband will be devoted to 5G.

Other bands such as the L, S, C and Xband are more common to applicationswith less important need in throughputsuch as M2M. It is also commonly usedby earth observation LEO projects.

V bands, the future for nextgeneration Very High ThroughputSystems?

In June 2016, Boeing revealed its plan fora LEO constellation made up of 2956satellites operating at 1200km in the Vband. This band is currently quasiunused because of technical difficultiesassociated with its high position in thespectrum.

However, this band has several meritsbecause of the large amount ofbandwidth available. There should notbe thus any sharing issue similar to theone experienced in the C and Ka bandtoday, especially for what pertains to therepartition of the usage betweenterrestrial and satellite players.

Since Boeing announcement, severalother projects using this band havesurfaced and been filled to the ITU. Thisinterest is not limited to LEO projects butalso shared by MEO and GEO players.

In many ways, the V band is today in thesame position as the Ka band in the1990s. It has been identified aspromising but propagationcharacteristics still need to be furtherstudied before commercial equipmentcan be developed. Those developmentscould take 10 years, before equipmentare available at an economical pricepoint.

References(1)Communication satellites by William Stalling

(2) High Throughput Satellite CommunicationsSystems: MEO vs. LEO vs. GEO(http://www.harriscaprock.com/blog/high-throughput-satellite-communications-systems-meo-vs-leo-vs-geo/)

(3) http://www.reuters.com/article/us-space-spacex-launch-idUSKBN1711JY

(4) http://spacenews.com/eutelsat-does-the-math-on-reducing-future-satellite-costs-by-at-least-20-percent/

3

Spectrum is key toincreasethroughputs butlimited inavailability

About the Digital Transformation MonitorThe Digital Transformation Monitor aims to foster the knowledge base on the state of play and evolution of digital transformation inEurope. The site provides a monitoring mechanism to examine key trends in digital transformation. It offers a unique insight intostatistics and initiatives to support digital transformation, as well as reports on key industrial and technological opportunities,challenges and policy initiatives related to digital transformation.

Web page: https://ec.europa.eu/growth/tools-databases/dem/

This report was prepared for the European Commission, Directorate-General Internal Market, Industry, Entrepreneurship and SMEs;Directorate F: Innovation and Advanced Manufacturing; Unit F/3 KETs, Digital Manufacturing and Interoperability by the consortiumcomposed of PwC, CARSA, IDATE and ESN, under the contract Digital Entrepreneurship Monitor (EASME/COSME/2014/004)

Authors: Vincent Bonneau & Basile Carle, IDATE; Laurent Probst & Bertrand Pedersen, PwC

DISCLAIMER – The information and views set out in this publication are those of the author(s) and should not be considered as theofficial opinions or statements of the European Commission. The Commission does not guarantee the accuracy of the data included inthis publication. Neither the Commission nor any person acting on the Commission’s behalf may be held responsible for the use whichmight be made of the information contained in this publication. © 2017 – European Union. All rights reserved.