Conference PaperPDF Available

Global coverage of multi-hop free-space optical ground-to-airliner data links

Authors:

Abstract and Figures

We present a model scenario in which airports and commercial aircraft are equipped with optical transceivers for high-speed onboard internet access. In a first step, basic line-of-sight calculations for ground station-to-flight and flight-to-flight are performed using timetable data published by major airline alliances. We then choose a set of relevant markets / arenas (both inter-continental and continental) to calculate respective aircraft reachability statistics. Based on this, the feasibility of single-hop and multi-hop data relays is estimated for typical scenarios. As several simplifying assumptions are made, we compare our data to actual historical flight data to judge the model's accuracy. Finally, we discuss major technical and commercial aspects for an eventual implementation of this dynamic free-space optical network.
Content may be subject to copyright.
978-1-4799-4891-8/14/$31.00 ©2014 IEEE
C4-1
GLOBAL COVERAGE OF MULTI-HOP FREE-SPACE OPTICAL GROUND-
TO-AIRLINER DATA LINKS
Alexander Wirthmüller, Laboratoire Temps-Fréquence, Université de Neuchâtel, Switzerland
Stefan Kalchmair, Laboratory for Nanoscale Optics, Harvard University, Cambridge, MA
Abstract
We present a model scenario in which airports
and commercial aircraft are equipped with optical
transceivers for high-speed onboard internet access.
In a first step, basic line-of-sight calculations for
ground station-to-flight and flight-to-flight are
performed using timetable data published by major
airline alliances. We then choose a set of relevant
markets / arenas (both inter-continental and
continental) to calculate respective aircraft
reachability statistics. Based on this, the feasibility of
single-hop and multi-hop data relays is estimated for
typical scenarios. As several simplifying assumptions
are made, we compare our data to actual historical
flight data to judge the model’s accuracy. Finally, we
discuss major technical and commercial aspects for
an eventual implementation of this dynamic free-
space optical network.
Introduction
The wish to provide passengers on
intercontinental flights with reliable high-speed
internet access calls for the evaluation of concepts
which go beyond solutions currently in use. Some
airlines offer access to ground-based GSM/3G
infrastructure, which however is limited to inhabited
regions. Other carriers equip their long-haul fleet
with radio frequency (RF) transceivers that connect
to a satellite-based system. While the latter offers
global coverage, it is hampered by bandwidth shared
between many customers, the impracticality of
maintenance and risk of eavesdropping due to the use
of radio signals with little directionality.
Ever increasing commercial air traffic and
recent successful demonstrations of optical free-space
ground-to-air data links give rise to the question
whether multi-hop optical communication can be a
solution for future high-speed internet access on
flights over remote areas. We try to answer this
question by analyzing recent timetable information
from major airline alliances along with historical
track points obtained from a commercial service.
This work is based on line-of-sight calculations:
atmospheric perturbations (due to weather and
temperature gradients) impeding communication are
taken into account only in so far as that the line-of-
sight distances obtained geometrically are reduced by
applying conservative assumptions. The extent to
which communication is affected by atmospheric
effects is highly dependent on the carrier wavelength
used. Due to low absorption and reduced Rayleigh
scattering, we think that the mid-infrared spectral
region (wavelengths of 8 µm and longer) can be
highly interesting from an implementation point of
view.
A dedicated software tool, BeamRelay [1], was
developed over the course of the project. The
necessity for this solution arises from the need to
process large amounts of data (GB range), handle
many calculations in parallel (cloud computing), and
to be able to store intermediate results (database). To
avoid overhead on writing code for other things (e.g.
database access code, user interface) than problem
specific functionality, a model-based approach with
automated code generation was chosen. The
functionality of BeamRelay is depicted in Figure 1.
Figure 1. BeamRelay Software Tool
Data Model and Algorithms
Our model relies on timetable data only for
flight density estimation. The simple calculations
C4-2
presented below are possible due to the key
assumptions that:
earth is a sphere of radius 6371 km. Actual
ground station elevations are added by
using NOAA’s global relief model [2].
Terrain impeding line-of-sight connections
is not considered specifically, rather a
general ground clearance requirement of
2000 m is used
the set of ground stations comprises all
1446 airports included in the timetables
used. They are assumed to have visibility
cones based at their elevations above mean
sea level (MSL) with a 85° half opening
angle. We think this large angle is
justified, as in case a ground station would
have terrain obstacles (e.g. lie in a valley),
for the local implementation an exposed
location in proximity would be found
flights pass along great circles at a
constant altitude of 10.5 km with constant
velocity. The assumed altitude is a lower
bound for typical cruise altitudes of long
distance flights. We discuss the impact of
climb / descent and of the constant
velocity assumption later on
Data Sources
The two primary sources used to obtain flight
information include:
star2011-12: the consolidated StarAlliance
timetable, valid December 13, 2011
through February 26, 2012. This period is
of interest as it includes low-traffic days
around the Christmas and New Year’s
holidays. Member airlines include A3, BD,
JK, JP, KF, LH, LO, LX, OS, OU, SK,
SN, TK and TP in Europe, AC, CO, JJ,
PZ, UA and US in the Americas, CA, NH,
OZ, SQ and TG in Asia, ET, MS and SA
in Africa and NZ in Australia/Oceania
owMar2014: the consolidated oneworld
timetable, valid March 7, 2014 through
April 4, 2014. Member airlines include
AB, AY, BA, HG, IB and S7 in Europe,
4M, AA, LA and XL in the Americas, CX,
EG, JC, JL, KA, MH, NU, QR and RJ in
Asia, and QF in Australia/Oceania
As of 2014, StarAlliance does not provide
consolidated timetables anymore. To be able to
calculate approximate joint StarAlliance/oneworld
flight density statistics, we define a timetable
star2011-12 which is the timetable star2011-12
shifted in time, so that it overlaps with owMar2014
while maintaining a weekday match. It is important
to note that the sets of airlines (and thus flight
numbers) in the two original timetables are disjoint,
as recent alliance changes (JJ and joint AA/US from
StarAlliance to oneworld) are not taken into account
in owMar2014.
Eventual multi-stop or connecting flights were
broken into their constituent nonstop flights. As
timetables state local departure / arrival times,
manual association of airports with regions
(continents / countries / states) at a granularity
allowing for the distinct specification of the
respective location’s time zone had to be performed.
Finally, aircraft types were classified into wide
body / narrow body / regional categories and standard
seating configurations were obtained from public
internet resources.
Data Validation
Data integrity was assured by a systematic
search for implausible average flight velocities (by
route), and by comparing the average flight velocities
on both directions of each pair of airports.
Basic Geometry Considerations
Table 1. Geometry Symbols
Symbol
Quantity
re
earth radius / MSL
af
flight path altitude above MSL
e
h
required horizon clearance above
MSL
e
g
ground station elevation above
MSL
d
f-g
flight-to-ground station great circle
distance at af
d
f1-f2
flight-to-flight great circle distance
at af
α
g
ground station-based cone half
opening angle
α
'
g
α
g
equivalent cone half opening
angle at a
f
α
f
flight visibility cone half opening
C4-3
angle
angular geo-coordinate
normalized cartesian geo-
coordinate
angular coordinate in flight path /
communication corridor coordinate
system
normalized cartesian coordinate in
flight path coordinate system
Ā
rotation matrix such that to
transform geo-coordinates into
flight path / communication
corridor coordinates
flight path / communication
corridor angular distance
ω
flight angular velocity
The key measures for the feasibility of relay
connections are the possible flight-to-ground station
df-g and flight-to-flight df1-f2 distances covered, as
depicted in Figure 2.
Figure 2. Basic Line-of-Sight Geometry
The assumption for df1-f2 is that no connection is
made beyond the horizon, and that above the horizon
a margin of eh is maintained. Thus, the distance
theoretically possible for two aircraft cruising at the
same af is halved - this is done as a tribute to
atmospheric scattering.
The relevant geometrical relations read
(
(1)
(
(2)
(
(3)
(
(4)
where the angles αg = 85° and αf = 2.958° are
introduced to define the ground station and flight
path visibility cones, respectively. As af and eh are
fixed, αf is valid universally. The solution for (1)-(3)
needs to be found numerically.
For many calculations in BeamRelay, it is useful
to determine a flight path specific coordinate system
with rotation matrix Ā in which the flight path
between begin location rb and end location re lies in
the equatorial plane as
(
(5)
(
(6)
where is the flight path angular distance.
Mutual Visibility Calculation
The computation effort for mutual visibilities is
reduced significantly by first considering the static
constellation of flight paths and ground stations with
respect to one another. Based on this data, time-
dependent calculations (i.e. the flights taking place on
the flight paths) are performed only when needed for
a specific connection limited in time.
Resulting line-of-sight distances are shown in
Figure 3.
C4-4
Figure 3. Theoretical Line-of-Sight Distances
Flight Path-to-Ground Station
For a pair flight path / ground station, the ground
station location is rotated into flight path coordinates
, and the equivalent cone half opening angle
α
'g is
calculated. In case of mutual visibility, two
intersections of the flight path with the equivalent
cone can be found analytically:
(
(7)
(
(8)
These intersections define the visibility interval
as depicted in Figure 4.
Figure 4. Flight Path / Ground Station Geometry
Figure 5 shows the area which can be covered
by ground stations / single-hop connections only.
Figure 5. Ground Station Line-of-Sight Coverage
for 𝜶𝒈 = 85° and 𝒂𝒇 = 10.5 km
Flight Path-to-Flight Path
The geometry of two crossing flight paths
(indices 1, 2) is shown in Figure 6 as the intersection
of the and planes. While all
possible lines-of-sight between flights are contained
within the region marked in red, the calculation is
limited to determining the points at which the origin
flight 2 enters and leaves the destination flight’s 1
visibility cones (and vice versa). For this purpose, the
origin flight, parameterized by , is rotated into
coordinates. This is followed by the
numerical solution of
(9)
(10)
for and representing begin and end positions
of the visibility in coordinates.
C4-5
Figure 6. Mutual flight path visibility
Communication Corridors
For practical reasons, the determination of relay
opportunities is limited to a band around the
connection’s great circle. The implementation is in
terms of communication corridors which are handled
in a similar fashion as flight paths, in particular by
defining visibility cones. Continental connections
consider a band 400 km in width (corresponding to a
cone half opening angle of αc = 1.799° while for
inter-continental connections 800 km are taken into
consideration (αc = 3.597°).
Flight-to-Ground Station
Due to the linear relation
between flight path angle and time, the calculation of
flight-to-ground station visibility is straightforward,
based on flight path-to-ground station visibility data.
Flight-to-Flight
The flight-to-flight calculation is more involved:
pre-filtering of flights to be matched includes the
evaluation of static mutual flight path visibilities, and
the analysis of the time spans during which the flights
are airborne. The actual duration for which two
flights can have line-of-sight contact can then be
determined by solving the dot product
(11)
numerically for t0 and t1, representing start and stop
times of mutual visibility.
Relay Calculation
Equipment Line-Up
In order to calculate a relay connection that
follows a specific direction (backward / forward)
along the communication corridor, it is necessary to
establish a linear equipment order ei, or line-up, for
the connection duration . Equipment
comprises both ground stations and flights ; the time-
dependent position i is defined by the respective
equipment’s coordinate within the
communication corridor. Instead of calculating
snapshots at defined intervals during the connection
time span, a list with insert, remove, to-before and to-
after operators is generated.
At first, all equipment available for the
connection is obtained from mutual visibility data
and an initial line-up at t0 is determined. Time stamp
calculation for insert and remove operators is
straightforward. For the to-before and to-after
operators, the time stamps at which equipment
switches position within the line-up need to be
determined. A simple toggle operator can not be used
due to the time t granularity of one second which can
yield multiple simultaneous operations involving the
same equipment. Mathematically, the problem
(12)
(13)
needs to be solved numerically for t, for flight-flight
position switches ; the index c denotes the
communication corridor and its coordinate system.
Flight-ground station position switches are calculated
in a similar fashion, with one of the angles
invariant in time.
Relaying
The relaying algorithm combines the equipment
line-up and the flight-to-ground station / flight-to-
flight visibility information. For a connection with an
(airborne) flight as the target, an attempt is made to
establish relays to one ground station ahead and one
ground station behind, within the equipment line-up,
at any time. The algorithm ensures that if a forward
C4-6
(backward) relay is possible, it is found. If no ground
station can be reached, a tree structure with all
reachable (flight) equipment is determined.
Forward (backward) relays are strictly forward
(backward) as far as flights are concerned. However,
to make a direct connection to a ground station, a
change in direction (forward / backward) is
permissible. At the time at which a forward relay is
established, the corresponding ground station end
point is located ahead of the target flight. As the
target flight advances, it is possible that it passes the
ground station that initially was the end point of the
forward relay. Although this effectively establishes a
second backward relay, it is maintained until
connectivity is lost. An exemplary relay situation is
shown in Figure 7.
Figure 7. One-Hop Backward Relay and Two-Hop Forward Relay for Flight LH401 (JFK-FRA) Close to
Nova Scotia on January 15, 2012 (star2011-12)
It has to be noted that for an actual
implementation the number of hops (the fewer the
better) and the relay duration (no need to re-connect
for long periods desirable) should be optimized.
A first step towards statistical analysis is taken
by determining the time evolution of the reachability
for each flight, independent of the specific relays
established. For selected flights, this is shown in
Figure 8: distinguished are one way (light green) and
both ways (green) connectivity to ground,
connectivity to at least one other airplane (yellow),
and no connection (red).
C4-7
Figure 8. Reachability for Some Long Distance Flights on January 15, 2012 (star2011-12)
Statistical Analysis
Arenas
Table 2 lists the world regions / arenas
investigated. They were chosen in order to achieve
geographical diversity and to take into account the
respective market shares of the concerned airline
alliances.
Table 2. Arena Definitions for Statistical Analysis
trans-oceanic arenas (“to”)
to1
North Atlantic arena: Europe vs. North
America
to2
South Atlantic arena: Europe vs.
{Argentina, Brazil, Chile, Uruguay}
to3
North Pacific arena: North America vs.
{East Asia, South East Asia}
to4
Continental U.S. vs. Hawaii arena
other inter-continental arenas (“ico”)
ico1
Trans African arena: Europe vs. South
Africa
ico2
Trans Siberian arena: Europe vs. East
Asia
ico3
Pan American arena: North America
vs. South America
continental arenas (“co”)
co1
European arena
co2
Contiguous U.S. arena
co3
China arena
Metrics
The local passenger densities and connectivity
success rates are defined as
(14)
(15)
in which the respective passenger counts are derived
from aircraft type seating capacities. For an entire
arena, the corresponding densities are integrated over
the element areas Ael, to obtain the lumped passenger-
hour densities and connectivity success rates, denoted
as D and P, respectively.
Initial Observations
For visualization, results were superimposed
with continent maps [3] in Mercator projection. The
approximate passenger densities in the various arenas
can be identified immediately as shown in Figure 9.
Striking features include the very high passenger
density above the North Atlantic and the simple
identification of airline hubs.
C4-8
Figure 9. Passenger Density by Arena
Furthermore, in Figure 10, the bundled daytime /
overnight operation of long-haul flights in major east-
west (to1, to3 westbound in blue) and north-south
(ico1, ico3 southbound in blue) markets is clearly
visible. It also becomes evident that most long-
distance routes are not overly weekday dependent.
Figure 10. Bundled Flights (star2011-12)
C4-9
Reachability by Arena
In both timetables, an average Wednesday was
chosen for analysis: February 8, 2012 for star2011-12
and March 12, 2014 for owMar2014. For initial
calculations, represented in the first and third result
columns of Table 3, all aircraft of the respective
alliance were assumed to be equipped with optical
transceivers. From the non-continental arenas, to1
stands out with the highest connectivity success rate.
The continental arenas feature almost full coverage
comparison with Figure 5 shows that this is not only
due to coverage with many ground stations but due to
many possibilities of multi-hop relays.
It should be noted that throughout all arenas, the
timetable with the higher D (star2011-12 in all cases
but ico3) also achieves the higher P.
Figure 11. Local Connectivity Success Rates
Table 3. Lumped Connectivity Success Rates by Arena
C4-10
Widebody Aircraft Only
With few exceptions, long-distance flights are
performed by widebody aircraft which constitute a
<10% subfleet of all commercial aircraft. Equipping
only them with optical transceivers was thus
investigated as shown in the result columns 2, 4 and 6
of Table 3. Figure 12 additionally reveals that the
added connectivity resulting when all aircraft are
equipped is provided close to land, where typical
short-haul flights take place.
Figure 12. Connectivity Success Rate All Aircraft
vs. Widebody Only (owMar2014)
Joint Staralliance / Oneworld Timetable
Motivated by the trend that a higher D results in
a higher P, we used a time-shifted version of
star2011-12 and superimposed it with owMar2014,
such that StarAlliance flights actually taking place on
February 8, 2012 take place on March 12, 2014. The
results can be found in columns 5 and 6 of Table 3.
Indeed, all arenas yield connectivity success rates
higher than both individual timetable’s P’s. A notable
exception is ico1 in which the combined P is higher
than the P of owMar2014 but lower than the P of
star2011-12. This can be attributed to routings which
are geographically separated in most parts.
Figure 13. Connectivity Success Rate owMar2014
vs. Joint Timetables
Model Improvements
There are two shortfalls of the model used. First,
flights on short sectors tend to be overrepresented
when assuming scheduled flight time. Second, actual
flight routing is affected by weather and air traffic
control constraints. The impact of both effects was
studied by using historical track points obtained from
[4] for owMar2014 flights in the week March 9, 2014
to March 15, 2014. The coverage of this data is
incomplete. While complete data sets exist for flights
above North America, Europe and Japan, some
flights are listed with their actual departure and
arrival times only ; for other flights, no detail
information could be obtained at all. A full analysis
using historical data only was thus not feasible, also
due to glitches in flight routing present in the data.
Velocity
The velocity discrepancy is a result of time
margins which are added in timetables to ensure
reliable operation. They take into account ground
operations and eventual delays. The time margin-to-
timetable time ratio is higher for short flights which
is nicely evidenced in Figure 14. It can be seen that
even actual flight velocities are reduced for flights
shorter than 2000 km, this can be attributed to
reduced speed in flights’ climb and approach phases.
C4-11
Figure 14. Velocity Discrepancy Between Actual
and Timetable Data
Altitude
The availability of flights for relay connections
is overestimated when assuming them to be present at
10.5 km even during climb and descent (cf. Figure 3,
lower part). Again, this effect is more pronounced for
short flights. An improved model could work with
altitude profiles as depicted in Figure 15: during
climb and descent, the flight has connectivity to its
origin and destination ground stations only and is not
available for additional relay connections. As short
flights may not reach af, a requirement is set up that a
flight path needs to be at cruise altitude for at least
1/3 of the great circle distance. The corresponding
calculations read:
(17)
(18)
in which is the resulting cruise altitude. For our
data, 1075 out of 10979 flight paths would have to be
reduced from 10.5 km altitude accordingly.
Figure 15. Improved Flight Altitude Profile
Routing
Finally, the historical waypoint data was used to
qualitatively compare great circle (model) routing to
actual routing. This analysis is of interest, as for high
connectivity success rates not only bundled operation
in time (cf. Figure 10) but also in space is
advantageous. Generally, traffic guided by ATC
waypoints and weather should thus improve the
model results. A comparison of flight densities is
shown in Figure 16. Non-great circle contributions
can be clearly identified.
Figure 16. Great Circle vs. Actual Routing
(owMar2014, 24h Period)
Implementation Aspects
In the following we argue for the
implementation of the proposed network by means of
optical free-space communication using the mid-
infrared spectral range. Before a commercial
realization could take place, numerous technical (and
C4-12
political) challenges would need to be overcome,
some of which are mentioned here.
Transceiver Hardware
Reliable communication between aircraft is
challenging, but can be achieved with existing
technology. Most systems available rely on RF links,
since the technology is well established. However,
they cannot provide sufficient data transmission
capacity for modern internet services.
Free-space optical communication offers various
advantages compared to RF solutions [5], mainly
higher data rates (1-40 Gbps vs. 200-800 Mbps) and
high directionality. Good directionality is desirable to
prevent interference and to increase communication
security. In addition, unwanted reflections from
foreign airborne objects are greatly reduced.
A common figure of merit for the performance
of free-space communication systems is the
maximum link distance at a given data rate and
emitter power. RF systems allow communication
over several hundred kilometers, however, at low
data rates only. While existing free-space optical data
links allow much higher data rates, distances are
limited to tens of kilometers [5]. The reason is
absorption and scattering of visible light in the
atmosphere. Furthermore, reasonable link quality is
achieved in good weather conditions only.
The dominant causes for atmospheric
perturbations of optical signals are absorption by
water vapor and Rayleigh scattering on dust particles.
Fortunately, at around 8 µm wavelength water vapor
is not absorbing, hence this region is called an
atmospheric transmission window. In addition,
Rayleigh scattering decreases with increasing
wavelength and becomes almost negligible around 8
µm wavelength. In this region the atmosphere is
highly transparent, making it perfectly suited for free-
space optical communication. Figure 17 compares
measured attenuation values for near-infrared and
mid-infrared data links in a real-world scenario for
various weather conditions [6].
110
0
2
4
6
8
10
12 Rain
Near-IR (1.345 um)
Near-IR (1.55 um)
Mid-IR (8.1 um)
Attenuation (dB/km)
Visibility (km)
Fog
Figure 17. Attenuation of Near-IR and Mid-IR
Light for Different Weather Conditions [6]
The essential components to implement a free-
space optical data link are a stable high power laser, a
fast detector and a precise tracking system.
Visible wavelengths are not well suited for free-
space communication, even though light sources and
detectors are highly efficient in this wavelength
range. Visible light is strongly scattered in the
atmosphere and already small laser powers can cause
severe damage to the human eye.
For near-infrared wavelengths (1.3 µm and 1.55
µm), technology is well-established and has been
pushed by the development of fiber integrated
communication networks. Erbium doped fiber lasers
provide high powers, stable emission and fast
modulation speed. Various photo-detector
technologies are available providing excellent
detectivity and detector response times. However,
due to atmospheric scattering, high data rates over
long distances can only be achieved using very high
laser powers [5].
The mid-infrared technology platform (8 µm
wavelength) on the other hand is less established, but
has matured sufficiently to be implemented in
commercial products. Quantum cascade lasers
(QCL’s) provide high power, good modulation speed
and are available for any wavelength from 3 µm to 20
µm. The most common photo-detectors for the mid-
infrared wavelength range are Mercury Cadmium
Telluride (MCT) detectors. Generally, these detectors
are slightly less efficient than their near-infrared
counterparts, but this is more than compensated by
the improved atmospheric transmission over long
distances [6]. Furthermore, mid-infrared light cannot
C4-13
penetrate aircraft windows and is eye-safe to
potential bystanders.
Tracking systems for the finding and dynamic
reconnection of partner aircraft certainly is a
challenge (moving origin and target). However, it is
done in military target tracking and should be
technologically feasible. Recently, a ground-to-air
tracking system has been demonstrated, which is
capable of streaming high-resolution video at 1 Gbps
from a flying Tornado jet aircraft moving at 800
km/h over a distance of 60 km [7].
Commercial Aspects
For the scenario presented here to work, all
aircraft involved would need to be equipped with
corresponding transceivers, which is a considerable
cost factor. In addition, the decision to implement the
proposed network would require the teaming up of
multiple leading airlines or even airline alliances.
However, the approach to equip widebody aircraft
only with transceivers may alleviate the initial
financial burden. As the widebody aircraft market is
held by the Airbus-Boeing duopoly, a strategic
alliance between those two companies might help
establishing our proposed technology as well.
It should be noted that the implementation cost
is to be compared to that of launching and operating a
network of satellites (with their known drawbacks),
so that we see our solution as a competitive
alternative.
Possible Extensions
With current air traffic densities, our
calculations (e.g. ico1, ico2) show little coverage
above sparsely populated areas. While little can be
done about lacking coverage above deep-sea areas,
connectivity e.g. above Siberia or close to a shoreline
could be enhanced by adding a set of captive balloons
at altitudes up to 4 km. No significant interference
with air traffic is to be expected as with no airport
in proximity – flight paths lie much higher in altitude.
Comparable systems have been put to use before [8].
In contrast to solar-powered airborne systems [9],
[10] which are unable to navigate against the wind
and have a limited power budget for optical
transceivers, a system of captive balloons could
operate autonomously by tapping into the solar or
geo-thermal power of a large surrounding area on the
ground.
Using our model, a communication corridor with
captive balloons could be established as shown in
Figure 18. Its width varies between wmin and 2 x df1-f2
at cruise altitude. The corresponding formulas read
(19)
(20)
Figure 18. Balloon Placement Along a
Communication Corridor
To give an example, for a balloon spacing of
160 km (altitude ab = 4 km) along the corridor, its
width varies between 576 km and 659 km, assuming
a cruise altitude of 10.5 km.
Conclusion and Outlook
In summary, we have presented an in-detail
study of global aircraft reachability above remote
areas by means of multi-hop free-space optical
communication. The software tool BeamRelay, which
was developed along the way allows for the fine-
grained specification of model parameters and is
easily extensible for advanced models or other data
sets.
It could be shown that already today, high
density markets such as the North Atlantic exist in
which success rates higher than 70% can be achieved.
Currently underway is the implementation of model
improvements which take into account more realistic
flight paths for short-haul flights. In addition, the
strategic placement of captive balloons above remote
(land) areas is studied in order to improve the results
for markets such as the Trans African or Trans
Siberian arenas.
We think that multi-hop free-space optical
communication is a promising solution for the future,
considering advancements in underlying technology
and ever-increasing air traffic. Mid-infrared lasers
C4-14
and detectors might turn out to be the key
technology, due to advantageous atmospheric
transmission and eye-safe operation.
References
[1] A. Wirthmüller, BeamRelay:
http://www.epsitechnologies.com/preview/styled/styl
ed-2/index.html, retrieved on March 29, 2014
[2] NOAA, ETOPO1 global relief model:
http://www.ngdc.noaa.gov/mgg/global/global.html,
retrieved on March 29, 2014
[3] NASA Goddard institute, Panoply:
http://www.giss.nasa.gov/tools/panoply/overlays/,
retrieved on March 29, 2014
[4] Flightaware, FlightXML service:
http://flightaware.com/commercial/flightxml/,
retrieved on March 29, 2014
[5] Y. Koishi et al., International Conference on
Space Optical Systems and Applications (ICSOS),
May 2011
[6] Paul Corrigan et al., Optics Express Vol. 17, Issue
6, pp. 4355-4359 (2009)
[7] ViaLight Communications GmbH VLC:
http://www.vialight.de/index.php?id=180, retrieved
on April 15, 2014
[8] Wikipedia article on Tethered Aerostat Radar
System:
http://en.wikipedia.org/wiki/Tethered_Aerostat_Rada
r_System, retrieved on April 20, 2014
[9] Google, Loon project:
http://www.google.com/loon/, retrieved on April 20,
2014
[10] BBC, Google’s acquisition of Titan Aerospace:
http://www.bbc.com/news/business-27029443,
retrieved on April 20, 2014
Acknowledgements
The authors would like to thank Lubos
Hvozdara, Cornel Stücheli, Ophir Gaathon and
Martin Hoffmann for useful discussions.
Email Addresses
awirthm@gmail.com
2014 Integrated Communications Navigation
and Surveillance (ICNS) Conference
April 8-10, 2014
... Lasercom technology's drawback of transmission weakening over long distances can also be improved using a multipath system (Wirthmuller and Kalchmair, 2014). Adding relay nodes and repeaters would ensure strength of the signals and perform better than a single beam. ...
Article
The satellite telecommunication industry is going through a period of great transition due to changing user demands, technology advancements and evolving markets. The incumbents are having to diversify into new areas such as broadband connectivity, high performance satellites and constellations. This Team Project discusses the current state of the satellite telecommunication industry including the legal, economic, technological and social factors impacting the industry. This is then followed by identifying and analyzing a number of emerging trends in the industry, including the shift towards low latency, high secure communications, increase in bandwidth demand, laser communications, frequency allocation challenges, and the like. Following this analysis, market opportunities are identified for both new entrants and incumbents to leverage upon, namely, autonomous vehicles, mobile broadband, secure communications, deep space communications networks and enhanced event broadcasting technologies. Finally, recommendations are provided for approaching these opportunities using strategic tools and technologies, such that the satellite telecommunications industry can stay relevant in the era of 5G, Internet of Things and other upcoming developments.
Conference Paper
Full-text available
Deep Space communication systems are integral to modern space weather prediction, science missions and exploration. Historically these undertakings have been in the milieu of governments and government-based agencies. However, the emergence of NewSpace presents a challenge to this paradigm. Current Deep Space communication systems in the United States and elsewhere are aging and in need of additional throughput capacity. The responsibility for these upgrades now lies with more players than just national space agencies, as more entrants emerge in deep space including private entities and universities. Not only are there numerous planned scientific exploration missions, but there exist emerging concepts such as a permanent human presence in deep space, and commercially driven enterprises such as asteroid mining that will require the support of robust communications. The urgency of such an expansion and upgrade to the network is emphasized by analyzing the space telecommunications market and the supporting infrastructure, along with opinions from industry and professionals alike. This paper presents the argument for a commercial Deep Space Network (DSN) to support future commercial and government missions. Options for the formulation, organization and profitability of such a system are discussed. Next, private company undertakings, public private partnerships (PPPs) and international cooperation methods are examined. Finally, recommendations for the next steps are outlined.
Article
Full-text available
Mid-infrared (MIR) free space optical communication has seen renewed interest in recent years due to advances in quantum cascade lasers. We present data from a multi-wavelength test-bed operated in the New York metropolitan area under realistic weather conditions. We show that a mid-infrared source (8.1 microm) provides enhanced link stability with 2x to 3x greater transmission over near infrared wavelengths (1.3 microm & 1.5 microm) during fog formation and up to 10x after a short scavenging rain event where fog developed and visibility reduced to approximately 1 km. We attribute the improvement to less Mie scattering at longer wavelengths. We confirm that this result is generally consistent with the empirical benchmark Kruse model at visibilities above 2.5 km, but towards the 1 km eye-seeing limit we measured the equivalent MIR visibility to be > 10 km.
Panoply: http://www. giss. nasa. gov
  • Nasa Goddard Institute
  • Paul Corrigan
Paul Corrigan et al., Optics Express Vol. 17, Issue 6, pp. 4355-4359 (2009)
Wikipedia article on Tethered Aerostat Radar System: http://en.wikipedia.org/wiki
  • Vlc Vialight Communications Gmbh
ViaLight Communications GmbH VLC: http://www.vialight.de/index.php?id=180, retrieved on April 15, 2014 [8] Wikipedia article on Tethered Aerostat Radar System: http://en.wikipedia.org/wiki/Tethered_Aerostat_Rada r_System, retrieved on April 20, 2014 [9] Google, Loon project: http://www.google.com/loon/, retrieved on April 20, 2014
Google's acquisition of Titan Aerospace: http://www.bbc.com/news/business-27029443
BBC, Google's acquisition of Titan Aerospace: http://www.bbc.com/news/business-27029443, retrieved on April 20, 2014
BeamRelay: http://www. epsitechnologies. com/preview/styled
  • A Wirthmüller