Rural 5G Small Cells: Connecting the Unconnected Today and Tomorrow
CONNECTING THE UNCONNECTED THROUGH MEGACONSTELLATIONS · CONNECTING THE UNCONNECTED THROUGH...
Transcript of CONNECTING THE UNCONNECTED THROUGH MEGACONSTELLATIONS · CONNECTING THE UNCONNECTED THROUGH...
CONNECTING THE UNCONNECTED THROUGH MEGACONSTELLATIONS
TELECOMS WORLD MIDDLE EAST 2018
DUBAI - 18 SEPTEMBER 2018
Christian von der Ropp
Christian von der RoppStuttgart, GermanyIndependent Telecoms Consultant
Director Terrestrial NetworksAdvising stakeholders to
connect unserved and
underserved communities.
About
economic opportunity better education better healthcare efficient disaster relief
At AHUMANRIGHT we believe that access to information, and the Internet, should be a human right.
We spread that message, and work to connect the disconnected people.
Internet Access a Human Right?
Challenges of Geostationary Orbit
Today 99% of global satellite communication
runs via satellites in geostationary orbit (GEO)
High launch costs due toenergy required to reach GEO
(~58MJ/kg)
High latency due to signaltravelling 2×35,786km
RTD >600ms
Signal strength required tobridge distance dictates heavy
satellites with large solar arrays
(average mass 2016: >4000kg)
Limited Coverage by GEO Satellites
● Power budget constraints of GEO satellites paired with the great
transmit power required to bridge the distance to Earth means
spot beams must be limited to certain areas
● Satellite operators focus coverage to more densely populated
areas with higher bandwidth demand and revenue opportunities,
leaving those most in need under- or even unserved.
Internet Adoption 2016
© World Bank
835 million peoplein Africa
2.2 billion peoplein Asia
3.6 billion people lack Internet access
● GEO satellites operating in the same
frequency band (virtually unavoidable)
must observe a spacing of 2 degrees
(1470km of orbital extent per slot) to
avoid interference
● Obtaining rights to an orbital slot to
accommodate a new satellite over
populated regions is impossible
● Capacity can only be increased by
replacing older satellites by more
efficient ones as all relevant slots are
taken
GEO Belt is Congested
Why GEO? Basics of Orbital Mechanics
● Satellites are subject to gravitational pull which
must be overcome to avoid re-entry
● Propulsion not viable due to amount of
required fuel reserves (usual design lifetime of
GEO satellites is 15 years)
● Simple solution is making use of inertia - the
resistance, of any physical object, to any
change in its velocity.
● Give impulse to satellite (by launch vehicle)
● A balance between satellite’s velocity and
gravitational pull results in a stable orbit
Why GEO? Basics of Radio Communications
● In terrestrial radio communications signals can
be reflected and scattered by objects such as
buildings or the topography – a direct line of
sight between transmitter and receiver is not
necessarily required
● For a satellite link however a direct
unobstructed line of sight is required due to
the high frequency bands used for broadband
communications paired with the low signal
strength resulting from attenuation in
atmosphere
e ec on ca e in
l at o a a on
nobst cted
ine o i t
Difference between satellite-based andterrestrial radio communications
Why GEO? Basics of Radio Communications
● Because of the low signal strength parabolic
reflectors are required which bundle the signal
and focus it on the feed horn
● Imagine what would happen if the satellite
would be moving constantly…
Why GEO? Because it Doesn‘t Move (seemingly)
● Due to the impracticability of constantly adjusting
satellite dishes and out of the need to permanently
receive a signal satellites have traditionally been put
in an o bit w e e t ey move sync ono s to Ea t ’s
daily rotation.
● An orbit where a satellite circles Earth
synchronously with the day-night-rotation is called
geosynchronous or geostationary orbit (simplified).
● A geostationary satellite maintains a fixed position
in the sky as observed from Earth.
● Dictated by the laws of physics a satellite in
geostationary orbit must be located at an altitude of
35,786km and travel with an orbital velocity of
3.07km/s or 11,600km/h.
35,786km
3.07km/s
Implications for Lower Orbits
● Geostationary satellites are located at an altitude of
35,786km and travel with an orbital velocity of
3.07km/s or 11,600km/h.
● Satellites in lower orbits are exposed to stronger
gravitational pull and hence need a higher orbital
velocity to offset the earlier
● The lower the orbit, the higher the speed, the less
time required to circle Earth
● = Kepler’s Second Law:
● A line joining a planet [satellite] and the Sun [Earth]
sweeps out equal areas during equal intervals of
time.
GEO vs. LEO
Using satellites in Low-Earth Orbit would yield
many benefits over Geostationary Orbit…
Lower launch costs due tolower altitude and less cosmic
radiation
Low latency due to signaltravelling 1/30 of distance
RTD ~50ms vs >600ms
Shoter distance means weakersignal required, hence lesspower consumption and
smaller satellites(~150kg)
GEO vs. LEO
…but create numerous issues
Satellites would be movingconstantly at high speed
(100min per orbit, visible foronly 10min)
= many satellites and trackingantennas required
Each satellite could only covera fraction of Ea t ‘s surface
= many satellites and numerous gateways required
Satellites would causeinterference to geostationarysatellites while crossing the
equatorial belt= interference mitigation
First Proposals for LEO Constellations in the 1990s
● In 1994 Teledesic started working on a LEO constellation of 840 satellites providing 720Mbps each
● Backed by cellular pioneer Craig McCaw, Bill Gates and Saudi prince Alwaleed bin Talal
● Cost of $9bn and dot-com crash in 2000 led to failure of the project
…some LEO Constellations materialized though
● Iridium (1998) and Globalstar (1999) both launched LEO constellations which are still operational
● However the low frequency bands used (L- & S-band) limited offering to narrow band communications (<10kbps)
● Both filed for bankruptcy (Iridium in 1999, Globalstar in 2004) and Globalstar is still financially troubled
● Emergence of private space enterprise and competition means satellites are much cheaper to build and
launch (SpaceX launch costs ~ $65m vs. A‘s $422m price tag)
● Continued price pressure coming from new ventures entering the launch market (Rocket Lab, Blue Origin,
Virgin Galactic etc.)
● Technological advancement has led to miniaturization of satellites (cubesats, small satellites) resulting in
lower size, weight and costs for each satellite
● Cellular technology can be adapted to handle the complexity of the thousands of spotbeams and the
frequent handovers of millions of terminals between the same
● Efficient line production of satellites instead of manual assembly in laboratory environment
● Market has grown massively as Internet access is considered a necessity today
What has changed since the 1990s?
The Participants in the Megaconstellation Race
„STARLINK“
…and potentially further entrants such as Boeing (perhaps with A le‘s backing?) or Facebook („At ena“ project) and especially from China or perhaps a Chinese-Russian JV is expected
● Very high number of satellites (>100)
● global coverage
● including polar coverage
● very high capacity (multi-Gbps)
● High frequency re-use as in High-
Throughput Satellites
● low latency (<50ms)
● technological complexity
● new manufacturing methods
Defintion of „Megaconstellations“
Connect everbody and everything
● Cellular backhaul
● Aircraft
● Ships
● Oil rigs
● Mining facilities
● Emergency vehicles
● Schools, hospitals and public buildings in the
developing world
● Enterprises (especially LeoSat)
Market for Megaconstellations
Overview of Megaconstellations
„STARLINK“
Number of satellites 882 (+2,000) 112, up to 512 78-108 4,425
Altitude 1,200km (+8,000km) 1,000km/1,248km 1,430km 1,200km
Capacity per satellite(fwd only)
7.2Gbps n/a 1.6-5.2Gbps 10Gbps
First prototype launch
late 2018/early 2019 12 January 2018 no protoypes planned 22 February 2018
Completion of initial constellation
2022 2022 2022 2022
Beam design fixed (global) steerable spotbeams steerable spotbeams fixed (global)
Frequency band (userbeams)
Ku (+Ka, V & E) Ka Ka Ku+Ka
Overview of Megaconstellations
„STARLINK“
Cost per satellite $700k-$900k n/a n/a n/a
Cost of initial constellation
$6bn n/a $3.6bn $10bn
Equity $1.7bn in equity n/a $100m (Series A) n/a
Other Funding Remainder to befilled by Coface (French export bank)
n/a $1bn in revenuecommtiments
n/a
Backers Softbank, Airbus, MDA, Bharti, GrupoSalinas, Intelsat, Hughes, Coca-Cola, Qualcomm, Virgin
n/a JSAT, Hispasat n/a
Small Satellites vs. GEO Satellites
Dimensions 1m³ ~ 100m³
Weight 150-200kg >4,000kg
Design Liftime 5-7 years 15+ years
Launch cost ~$ 250,000 >$ 50,000,000
Cost per satellite $500,000-$700,000 $290,000,000 (typically)
● OneWeb Satellites (JV with Airbus) is building
$85 million factory in Exploration Park, Florida
enabling efficient line production of small
satellites
● Peak production capacity of 15 satellites per
week targeted
● Near the Kennedy Space Center, Florida
(SpaceX and ULA launch facilities)
● Contract manufacturing for third parties
planned
Mass Production of Satellites
● Launching hundreds or even thousands of
satellites will be a massive undertaking
requiring dozens of separate launches
● Shorter design lifetime will require frequent
replacments of satellites starting just a few
years after completion of initial constellation
● SpaceX as launch provider has the advantage of
access to own rockets
● OneWeb alone will need at least 25 Russian
Soyuz-11 rockets to launch every 3 week for
the first 882 satellites (36 per launch)
Launch Capacity & Satellite Replacements
● Due to the constant movement (relative to Earth) of LEO satellites tracking
antennas are imperative
● Such will have to track the serving satellite and establish a link to the next
satellite before the earlier goes out of range in order to enable gapless hand
overs for continous connectivity
● Existing mechanical tracking antennas are large (unsuitable for many
applications such as moving vehicles) and prohibitively expensive (~ $200k)
● Flat panel antennas based on phased-array technology and metamaterials
are under development by:
● Kymeta (JV with Sharp for production on LCD TV lines)
● Phasor
● Isotropic Systems (founded by former O3b staff)
● Alcana Systems (backed by Merck AG, Germany)
● Performance and interference (side-lobing) still a problem
● Will they reach a price point allowing mass adoption?
Tracking Antennas
Inter-Satellite Links
Satellites too far from nextgateway cannot provide service
Without ISLs
• OneWeb‘s first generationof satellites will only be ableto provide service while in line-of-sight with a gateway
• Once outside a ateway‘sreach satellites won‘tprovide coverage
With ISLs
• Satellites interconnected byISLs can keep forwardingtraffic to adjacent satellitesuntil a satellite with a gateway link is reached
• One satellite connected to a gateway could support entire constellation(although with a bandwidthbottleneck and a singlepoint of failure)
Traffic is routed in orbit to a satellite with gateway link
● OneWeb will require 50-60 gateways globally to
enable global coverage
● Constellations with ISLs will – in theory - need less
gateways because traffic can be routed in orbit.
● Especially LeoSat highlight the capability to route
traffic from door to door without touching the ground
– independence from terrestrial infrastructure means
less vunerability to outages and interception
● However many regulators will require operation of
national gateways for control and intercept purposes
● Latency requirements of MNOs and enterprises (cloud)
might require more regional gateways than anticipated
Implications for Ground Segment
● Except for OneWeb all Megaconstellations will support
inter-satellite links using optical laser links and allow in-
orbit routing
● speed of light
● in vacuum: 299,792,458 m/s
● In fibre: 204,190,477 m/s
● traffic routed in low-Earth orbit using inter-satellite links
will have substantial latency advantage over meandering
submarine cables on most intercontinental routes
● New York - Tokyo: 100ms vs. 135ms
● HFT will switch to satellite
Inter-connected Megaconstellations will cut latency
● ITU assigns orbital slots but operators are still
subject to approval of every single national
regulator allowing the use on the ground
● ITU framework is designed for GEO and has
no provisions for spectrum sharing
● Instead frequency rights are assigned on first-
come first-served basis; whoever launches
first would block huge amounts of spectrum
● Any later market entrant is obliged to avoid
interference – extremely challenging as
coverage would be impaired
Regulatory Issues
● Market access to some nations will be difficult
to impossible (Russia already denied approval
to OneWeb)
● Governments might oppose architecture
(independence from terrestrial infrastructure)
requiring ability to intercept, filter or have a
kill-switch to disable service (e.g. Iraq which
disables Internet during school exams period)
● FCC requires any Megaconstellation to launch
half of constellation within six years from
approval or license might be revoked
Hystery
Risk of Interference with GEO
While crossing the equatorial belt LEO satellites can potentially interfere with GEO
satellites; Mitigation technologies and potentially fall back to GEO required;
● The Kessler syndrome (also called the Kessler effect of
collisional cascading), proposed by the NASA scientist Donald J.
Kessler
● Objects in orbit travel with extremely high speeds relative to
each other (several tens of thousands km/h) – impacts fatal!
● It is a scenario in which the density of objects in low-Earth orbit
(LEO) is high enough that collisions between objects could
cause a cascade where each collision generates space debris
that increases the likelihood of further collisions
● One implication is that the distribution of debris in orbit could
render space activities and the use of satellites in specific
orbital ranges infeasible for many generations
Fear from the Kessler Syndrome
● Euroconsult expects about 7,000 new smallsats to be
launched over the next 10 years,
● six-fold increase from the 1,200 units launched over the
past decade
● Smallsat-dedicated launch services are expected to
generate $16 billion during the next ten years.
● Launch and manufacturing of the 7,000 smallsats valued
at $38 billion = quintupling decade-to-decade.
● By 2022, an average of 580 smallsats will be launched
every year
● This compares to an annual average of 190 satellites
launched over the past five years
● Average to jump to 850 satellites per year on subsequent
years up to 2027 because of the deployment of one mega
constellation
Low-Earth Orbit to be Populated
● Collision avoidance will be of greatest importance to avoid Kessler Syndrome
● Active collision avoidance will require propulsion which currently is not planned
on any Megaconstellation
● Though OneWeb plans drag device (kind of sail) that can initiate and accelerate
deorbiting process after lifetime ends
● Passive collission avoidance to be enabled by LeoLabs, a company planning a
ground-based radar system tracking objects in orbit (raised $13m in July 2018,
backed by Airbus)
● Current deorbit requirement for satellite operators in the US is 25 year post-
mission or 30 year after launch – too long many industry experts say
Space Debris Mitigation
● Northern Sky Research (NSR) expects that anywhere from
30 to ~5,000 LEO satellites can possibly launch in the next
decade.
● NSR further estimates that if all constellations do go ahead
as planned, we are looking at capacity of the order of 20-
30 Terabits coming online in the next decade, which is
orders of magnitude higher than the 2.5 Terabits of
GEO/MEO HTS capacity that NSR expects by 2023.
● Even if the industry remains divided on whether all this
capacity constitutes a “bubble”, it remains feasible that 20-
30 Terabits of capacity could be launched by 2020.
● NSR would wager about 2 or 3 Megaconstellations
actually launch.
Industry Expectations
● GEO satellite orders are in sharp decline
● NSR: Non-GEO satellites to dominate supply, adding over 25
Tbps in next decade (18 September 2018)
● Maxar considering shuting down GEO business (July 2018)
● Boeing acquires Millennium Space Systems, a provider of
agile, small-satellite solution (July 2018)
● Industry shifting focus to small satellites to be launched in
LEO reflects confidence in Megaconstellations actually
materializing
● Defense and intelligence sector also preferring the agility and
lexibility o small satellites (e. . DA A’s Black Jack
programme)
Industry ExpectationsGeostationary satellites order per year
"I think this is a critical time in the history of humanity,
because for the first time people are going to be coming
together from every corner of the world and be able to
engage in direct conversation with each other. [...]
In the next 5, 10 to 20 years the rest of the world is going to
come online. The next two billion people are going to come
online. For the first time we are going to hear from people,
we have not heard from before."
Jimmy WalesWikipedia Founder
photo: © VGrigas (WMF) (CC-BY-SA-3.0)
THANK YOUFOR YOUR ATTENTION
Christian von der [email protected]
For frequent updates join my LinkedIn group:
LEO-HTS Megaconstellations
search on LinkedIn for „Megaconstellations“ or go to www.megaconstellations.com