See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/26571587

A WiMAX Payload for High Altitude Platform Experimental Trials

Article  in  EURASIP Journal on Wireless Communications and Networking · December 2008

DOI: 10.1155/2008/498517 · Source: DOAJ

CITATIONS READS

5 219

3 authors, including:

Tim Tozer
The University of York
152 PUBLICATIONS   1,845 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Tim Tozer on 16 September 2014.

The user has requested enhancement of the downloaded file.

Hindawi Publishing Corporation
EURASIP Journal on Wireless Communications and Networking
Volume 2008, Article ID 498517, 9 pages
doi:10.1155/2008/498517

Research Article
A WiMAX Payload for High Altitude Platform
Experimental Trials

John Thornton, Andrew D. White, and Tim C. Tozer

Department of Electronics, University of York, York, YO10 5DD, UK

Correspondence should be addressed to John Thornton, jt21@ohm.york.ac.uk

Received 27 September 2007; Revised 21 February 2008; Accepted 6 May 2008

Recommended by Ryu Miura

The Swiss-based “StratXX” project during 2006-2007 has been developing a stratospheric lighter-than-air platform for deployment
of telecommunications and environmental monitoring services. In support of a first round of experimental trials, a WiMAX
communications subsystem, including platform payload and ground stations, has been developed. The communications system
design and results of terrestrial pretrials are reported.

Copyright © 2008 John Thornton et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1. INTRODUCTION “Pathfinder Plus” craft: this demonstrated high definition

TV in UHF and a video connection using an off-the-shelf
Wireless local and metropolitan area networks (WLANs/ W-CDMA cell phone from a payload limited to 50 kg.
WMANs) continue to gain acceptance and commercial Most other activity has been only with low altitude craft
rollout, enabled typically through the IEEE 802.11 and .16 or has been supportive work such as direction-of-arrival
family of standards. While these are generally terrestrially- measurements. A later NICT trial with an airship in 2004
based networks and now technologically quite mature, a again demonstrated HDTV and also autonomous flight,
contrasting approach is that of the high altitude platform albeit at an altitude of 4 km. NICT have also developed a
(HAP) where very high system spectral efficiency could be range of steerable antenna solutions for 28 GHz, 31 GHZ and
obtained by use of cellular networks and spot beam antennas 48 GHz, but these have only been trailed at low altitude.
on a platform payload located at an altitude of around 21 km The European 5th Framework Helinet programme [5]
in the stratosphere [1–3]. conducted mainly theoretical studies, while the 6th Frame-
A number of programmes are promoting and now testing work CAPANINA project [6–8] succeeded in demonstrating
HAP communications, and a concept gathering momentum broadband communications at up to 11 Mbps using a “one-
is the deployment of the 802.16 family, also called Wireless shot” stratospheric balloon to 24 km altitude and 40 km
Interoperability for Microwave Access (WiMAX) to enable ground distance using an 802.11 basestation with frequency
HAP-terrestrial links. WiMAX is increasingly seen as the conversion to the 28 GHz carrier frequency and a high gain,
future basis for broadband wireless Internet services, and is tracking reflector antenna at the ground station (see “Trial 2”
well suited to delivery from HAPs, which would effectively in [9]).
represent very tall radio masts. WiMAX thus represents a The StratXX programme involves a number of trials,
logical choice for development from this type of platform, phased towards a full sized stratospheric airship demon-
and has been selected by the StratXX programme as offering stration representative of commercial service provision.
near-term commercial benefit in this context. Following their experience of developing and executing the
To date, only very limited practical trials of the com- CAPANINA payload and wireless trials, a team at the Uni-
munications capabilities of HAPs have been reported. Some versity of York has been tasked with developing the WiMAX
stratospheric communications trials were conducted in 2002 payload and communications subsystem for the first full
by CRL (now NICT) of Japan [4] with an Aerovironment trial of the StratXX platform. This paper reports briefly on
2 EURASIP Journal on Wireless Communications and Networking

Table 1: Alvarion’s BreezeMAX radio specifications.

HAP
payload
3450–3500, 3550–3600,
Radio frequency (MHz)
Uplink (UL) Downlink (DL)
Channel bandwidth 3.5 MHz/1.75 MHz
BS+28 dBm
Output power
CPE+20 dBm
WiMAX Modulation OFDM, 256 FFT points
Convolutional coding:
Forward error correction
rates 1/2, 2/3, 3/4

CPE1 CPE2 CPE3 WiFi

LAN Table 2: Alvarion’s BreezeMAX adaptive modulation levels.
ADSL/cable
Net bit rate Sensitivity
Modulation and coding “rate”
Internet PDA (Mbps) (−dBm)
BPSK 1/2 1 1.41 100
BPSK 3/4 2 2.12 98
Figure 1: Typical WiMAX-from-HAP trials scenario.
QPSK 1/2 3 2.82 97
QPSK 3/4 4 4.23 94
the requirements, constraints, system design, link budgets, 16QAM 1/2 5 5.64 91
choice of equipment, and so forth, and then presents some 16QAM 3/4 6 8.47 88
early results of communications link performance from 64QAM 2/3 7 11.29 83
terrestrial and airborne pretrials. 64QAM 3/4 8 12.71 82

2. SCENARIOS AND SYSTEM REQUIREMENTS

may be severe. An important factor is whether the payload
Our end goal was the demonstration of broadband wireless should or not be pressurised and it was decided early in the
services over a link length in excess of 20 km between design process that pressurisation should be used so that
an HAP and at least one ground station or customer off-the-shelf equipment could be used with minimal risk
premises equipment (CPE). The link should also support of it failing due to pressure loss. Various system options
data streaming from an onboard optical sub-payload. A were considered such as whether to modularise the payload
typical scenario is illustrated in Figure 1. subsystems into separate pressurised containers, whether to
The essential features of Figure 1 include use steered or nonsteered WiMAX antennas, whether to use
additional WiMAX channels for a dedicated backhaul and
(1) one CPE acting as the gateway to the wired internet; also whether this might use a dedicated millimetre-wave
(2) at least one CPE is “remote” and is linked to the wider link. Several of these options had to be abandoned due to
internet via the HAP payload; mass, cost, and also time constraints, and the current design
(3) further users might utilise a WiFi access point whose is summarised in Figure 2. Here, various details have been
backhaul is provided via the WiMAX CPE. omitted for clarity, such as the temperature and pressure
probes which feedback local environmental data to the
Numerous variants are of course possible. However, from the failsafe controller - this important bespoke item is in overall
typical scenario of Figure 1, certain inferences can already control of the payload and would selectively shut down
be highlighted. Since WiMAX is intended as a point-to- subsystems to control the thermal and power budgets as
multipoint WMAN, the base station must be placed on necessary. The other items in the payload can be summarised
the HAP so as to have in view all the CPEs. Furthermore, as
one of the CPEs must then be used, in effect, as the
backhaul station (CPE1 in Figure 1), which would tend to (1) PC servers—one to provide local web services and
act as a bottleneck when the system is heavily loaded. This some house-keeping functions and one to manage
architecture is therefore suitable as a minimal approach to the gimballed camera/video sub-system;
a first demonstration of WIMAX to the stratosphere, and (2) power conditioning unit (bespoke);
would benefit from a dedicated backhaul (possibly in another
radio band, e.g., 28/31 GHz) should the system be scaled-up (3) WiMAX sub-system (adapted);
with additional users. (4) telecommand and control.

3. PAYLOAD HIGH LEVEL SYSTEM DESIGN The need otherwise for a gimballed (steerable) payload spot
beam antenna was considered at some length before adopting
Any HAP communications payload is liable to constraints in a compromise configuration which allows switching between
terms of mass, volume, and available power, some of which relatively wide and narrow beams.
John Thornton et al. 3

Hard disk Hard disk Expendable

Flash storage cameras
Power Power IDE IDE
switching and
conditioning
unit Video
PC server 1 PC server 2 Camera
Video
RS232
Gimbal
Control
RS232
RS232 Ethernet
switch RS232-
TCP/IP
Fail safe
microcontroller
WiMAX base station
RS422

Power, control, Antenna 1

Cooling uplink & downlink IF
Tele command and fans
control RF
ODU switch
(Radio RF
modem)
Antenna 2

Figure 2: High-level payload system diagram.

4. WiMAX SUBSYSTEM and made ready for terrestrial field trials for purposes of
evaluating the WiMAX equipment prior to full HAP payload
Early in the design life cycle we carried out a survey of build. The antennas used at this stage were standard Alvarion
available WiMAX equipment and of suitable spectrum. The items, namely, a 14 dBi sector antenna and an 18 dBi CPE
remit was for use of bands close to 3.5 GHz for short- antenna, these being linear polarised and thus unsuitable for
term trials in Switzerland. Following a short evaluation the final HAP trial due to the polarisation misalignment that
phase, we chose Alvarion’s BreezeMAX micro-base station is likely to occur with changes in platform yaw angle.
and associated CPE equipment. The radio specifications for A temporary nonoperational radio license was obtained
these are summarised in Table 1. from Ofcom in the UK for operating the uplink at 3.485 GHz
Further specifications associated with the eight adaptive and downlink at 3.585 GHz. For this, the license application
modulation levels (also called “rates”) are shown in Table 2. detailed a number of sites in North Yorkshire, UK, from
These are specific to a 3.5 MHz bandwidth radio channel which a line-of-sight (LOS) to the University of York was
and quoted for a typical packet error rate of 1%. We should available (since the eventual HAP trial was for LOS, our
add that the BS can operate in frequency division duplexing interest lays in investigating the limit of performance for
(FDD) and time division duplexing (TDD) when working long range LOS links rather than non-LOS, high multipath
with more than one CPE, while each CPE can operate in conditions). A series of such trials was performed during the
TDD only. period March to August 2007. A typical mobile test arrange-
In hardware terms, the minimum BreezeMAX working ment is shown in Figure 3 where the WiMAX subpayload was
configuration comprises one BS, one BS outdoor unit (ODU, driven to the remote site in a van. The CPE was retained at
see Figure 2), BS antenna, one CPE outdoor (RF) unit, and the University of York. This arrangement allowed internet
CPE antenna. Ethernet ports at the CPE ODU and at the BS browsing at rates up to about 8 Mbps from remote sites at
provide network connectivity. At the CPE, a separate indoor distances of 30 km and beyond.
power supply provides −48 V DC to the ODU via power over During the trials the WiMAX system was driven by
ethernet cables. applications software so as to stimulate either the uplink
or the downlink to run at maximum possible data rates,
4.1. Terrestrial WiMAX evaluation for example, by demanding large file transfers of up to
70 MBytes. In parallel, various outputs from the operating
Hardware was procured for the above BreezeMAX config- system were recorded, chief among these are received signal
uration, along with operating system (OS) and computer strength (RSS), signal-to-noise-ratio (SNR), and “code rate”
servers. These were configured with applications software (at level 1 to 8, being, resp., the lowest and highest order
4 EURASIP Journal on Wireless Communications and Networking

rcirc

h
θsub

(a) Antennas on mast, Vale of York beyond

Figure 4: StratXX platform position keeping sphere.

time sampling bins are quite coarse, being about 30 seconds

(hence the coarse times logged under the first column) and
the logged data will be an average over each period. UL&DL
data rates within Table 3 are in effect TCP/IP throughputs
measured within the base-station network processor unit
whilst those described in Table 2 are “raw,” that is, before the
protocol overheads are removed. For example, a single CPE
with a DL of rate 8 which has a theoretical net bit rate of
(b) Payload with BS in vehicle 12.71 Mbps would achieve a measured TCP/IP throughput
of 9–9.5 Mbps.
Figure 3: A remote, portable WiMAX test rig. In parallel with preparation for these trials, a link
budget calculator was written for purposes of establishing a
theoretical link prediction tool against which measured data
could be compared (see Table 4). Here, several approaches
modulation levels shown in Table 2) for both UL and DL. are suitable, including a fully theoretical treatment based
An extract from many hundreds of lines of logged data is on required energy-per-bit-to-noise density; a simpler treat-
shown in Table 3 where the link distance is 20 km and the ment using the equipment’s sensitivity specification for each
combined antenna gain specifications are 32 dBi. Also a 20- modulation level; and a further empirical method based on
dB attenuator has been inserted in one of the RF antenna observed behaviour where we correlate logged “rate” against
interconnects so as to force the system to operate at rates SNR and received power. Following some refinement, these
below maximum, which to some extent emulates the effect approaches generally agree to within 2 dB or so. A typical
of a ten-fold increase in path length. such link budget is shown as a spreadsheet in Table 4 .
In Table 3, we would emphasise that the data is generated
by the WiMAX hardware and operating system, and we have
no direct way of measuring independently each parameter. 5. WiMAX ANTENNAS SUBSYSTEM
The absolute accuracy of this data should therefore be
regarded with a little scepticism: the nature of the logged Antennas are a particularly important component in any
data is intended for network administration purposes and wireless network and especially so when this is deployed by
should be interpreted with care if used in the context of an HAP. The platform displacement, pitch, roll, and yaw,
a more rigorous scientific analysis. Nevertheless it is worth along with carrier frequency have a profound influence on
pointing out some of the properties of the logged data. For the choice of antenna technology and its cost. Where a
example, although the link is nominally static (the terminals multicell layout is used to maximise spectrum reuse, the
are not moving), the logged SNR and RSS are not quite problem can be complex although tractable [10]. For present
static (nor are they logged to equal precision for UL and purposes, a single cell and hence beam only was needed, with
DL). Where the UL and DL throughputs show zero, this a ground footprint of around 12 km diameter. To mitigate
is most likely due to their not being stimulated, that is, against platform movement, a steered antenna was in many
no traffic is being demanded; demand first occurs in the ways preferred, although time and cost implications of either
second line at time 14:57 where the DL is stimulated until developing a bespoke gimbal or modifying a commercial
time 15:00. During this time, the UL indicates relatively little item were such that this approach was not pursued for the
throughput, and much of this traffic will be due to protocols initial trial. The station-keeping targets for the HAP are also
and acknowledgements. From 15:07 this scenario is reversed fairly stringent, these being a 5 km radius (rcirc ) position
and the UL is driven by demand. It is also evident that the sphere and maximum ± 5◦ pitch or roll, and from these it
John Thornton et al. 5

Table 3: Extract from logged WiMAX link data.

SNR UL SNR DL RSS, UL RSS, DL

Time Rate UL Rate DL Throughput UL (Mbps) Throughput DL (Mbps)
(dB) (dB) (dBm) (dBm)
14:56 10.7 17 −94.6 −87 4 6 0 0
14:57 10.1 16 −94.3 −88 4 6 0.121 5.416
14:58 10.4 16 −94 −88 4 6 0.11 4.918
14:58 10.9 16 −94.1 −88 4 6 0.119 5.406
14:59 11 16 −94 −88 4 6 0.109 4.934
15:00 10.5 16 −94 −88 4 6 0.129 5.91
15:00 10.3 17 −94.4 −88 4 6 0.128 5.915
15:01 10.7 17 −94.4 −87 4 6 0 0
15:01 10.3 16 −94.4 −87 4 6 0 0
15:02 10.4 16 −94.2 −88 4 6 0.129 5.952
15:03 9.9 17 −94.4 −88 4 6 0.13 5.908
15:03 10 16 −94.4 −88 3 6 0.07 3.279
15:04 10.7 17 −94.1 −88 4 6 0 0
15:05 10.5 16 −94.2 −88 4 6 0 0
15:05 10.3 16 −94.3 −88 4 6 0 0
15:06 10.7 16 −94.1 −88 4 6 0 0
15:07 10 16 −94.3 −88 2 6 0 0
15:07 10.2 17 −94.5 −88 4 6 2.657 0.054
15:08 10.2 16 −94.1 −88 4 6 2.699 0.056
15:08 10.2 16 −94.1 −88 4 6 2.647 0.061
15:09 10 16 −94.1 −88 4 6 2.652 0.06
15:10 11.1 16 −94.2 −88 4 6 2.698 0.057
15:10 9.9 16 −94.4 −88 4 6 2.38 0.054
15:11 10.2 16 −94.2 −88 4 6 2.696 0.06
15:12 10.6 16 −94.2 −88 4 6 2.646 0.061
15:12 10.3 16 −94.2 −88 4 6 2.447 0.055
15:13 10.5 16 −94.2 −88 4 6 2.698 0.056
15:13 10.6 16 −94.2 −88 4 6 2.653 0.061
15:14 10.4 16 −94.4 −88 3 6 2.655 0.063

is straightforward to derive optimum antenna beamwidth as that is, we solve for the maximum directivity D(θ) as a
follows. function of θHPBW at θ = (1/2)θt . (The approach is analogous
The HAP is intended to be on station at height h to that of optimising cell-edge directivity in [10].) From
directly over one of the ground terminals. Figure 4 shows this simple recipe we choose θHPBW = 32.6◦ (D = 15.8 dBi).
the geometry, where angle θsub is subtended by the position This represents the optimum HAP payload antenna HPBW
sphere. To this we add the range of pitch and roll change 2θ p given the above input data on station keeping. A narrower
to derive the effective total change in angle θt which must be beam will improve the link budget so long as the HAP is on
accommodated by the HAP antenna: station, but will degrade the link budget should the antenna
r pointing error meet the above limits. A similar treatment
θt = θsub + 2θ p = arcsin circ + 2θ p . (1) for the ground antenna, assumed fixed and so not subject
h
to pitch and roll error, yields an optimum beamwidth of
From which we derive θt = 39◦ . While a simple approach 24◦ (D = 18.5 dBi). For the purpose of procuring antennas,
might be to set the antenna half power beam width which tend to be specified in terms of gain, we have assumed
(HPBW) to this figure, we prefer to choose a slightly smaller an efficiency of 50% (this being quite reasonable for printed
beamwidth and one which maximises the gain at the angle of circuit array types which are commercially available), that is,
maximum displacement, that is, at θ = (1/2)θt . This may be G(dBi) = D(dBi) − 3 dBi. As a mitigation strategy against
expressed as the value of HPBW which satisfies the HAP station keeping error being excessive, a second
∂   antenna of lower directivity was added to the payload,
D(θ)θ=(1/2)θt = 0, (2) a gain of 7 dBi being here chosen. Circular polarisation
∂θHPBW
6 EURASIP Journal on Wireless Communications and Networking

Table 4: Form of typical link budget calculator.

Transmitter (CPE)
Power per carrier (dBm) 20.0
Antenna gain (dBi) 15.0
Antenna feed loss (dB) 1.0
EIRP (dBm) 34.0

Receiver (HAP)
The Boltzmann Constant (dBJ/K) −228.6
Noise Temperature (K) 300.0
Thermal noise density (dBm/Hz) −173.8
Receiver noise figure (dB) 3.0
Receiver noise density (dBm/Hz) −170.8
Antenna gain (dBi) 13.0
Cable loss at ground station 1.0

Modulation Scheme 64QAM 64QAM 16QAM 16QAM QPSK QPSK BPSK BPSK
Required Eb/No 16.5 16 12 11 9 7.8 8.2 8
Bit/symbol 6 6 4 4 2 2 1 1

Bandwidth (MHz) 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5
Code Rate 0.75 0.67 0.75 0.50 0.75 0.50 0.75 0.50
Data Rate (Mbit/s) (25% rolloff) 12.6 11.2 8.4 5.6 4.2 2.8 2.1 1.4
Data Rate (dBbit/s) 71.0 70.5 69.2 67.5 66.2 64.5 63.2 61.5
Required C/(Io+No) (dBHz) 87.5 86.5 81.2 78.5 75.2 72.3 71.4 69.5

input noise dBm = −108.39

Link Parameters P_received (dBm) −84.57 SNR 20.82
Frequency (GHz) 3.485
Wavelength (m) 0.086
Ground Distance (km) 5.0
Platform Height (km) 20.00
LOS Distance (km) 20.62
FSPL (dB) 129.6
Misc Atmospheric Losses (dB) 0.0
Edge of cell and antenna beam losses 1.0
Clear air losses (dB) 130.6

Alvarion sensitivity (dBm) −82 −83.0 −88.0 −91 −94 −97 −98.0 −100.0
margin derived from sensitivity −2.6 −1.6 3.4 6.4 9.4 12.4 13.4 15.4

Alvarion "code rate" 8 7 6 5 4 3 2 1

min SNR 21 19 16 12 9 7 5 3
margin derived from SNR vs rate −0.18 1.82 4.82 8.82 11.82 13.82 15.82 17.82

was specified. The two payload antennas are just visible in function of lateral displacement. This is shown in Figure 5 for
Figure 3(a) and also in Figure 9. The RF switch used is of h = 20 km and where we assume the ground station antenna
a coaxial electromechanical type and its measured loss does can be pointed (either manually or under electro-mechanical
not exceed 0.1 dB. The switching operation is controlled via control) to mitigate against occurrence of a further loss term
the telecommand system. here.
For the two payload antennas (13 dBi and 7 dBi) we then In Figure 5, the effect of the wider payload antenna beam
compute power rolloff due to pointing loss, using a simple is evident in allowing a link should the HAP drift beyond
curve-fit approximation [10], as a function of platform intended limits such as might occur during a launch phase. In
displacement. This loss can then be used in the link budget dimensioning the WiMAX link budget we have concentrated
calculator to derive expected modulation level (or “rate”) as a on the UL, since the CPE has available 8 dB less output power
John Thornton et al. 7

6 Upper bulkhead
Upper pressure cover
5
DC power GPS

4
Modulation level

Server control unit

Control/video/power
3 Base station
Central frame
2
IF
RF switch RF unit
1
RF cables
(Rate 0 = no link) Lower
0 pressure cover
5 10 15 20 25 Camera &
gimbal
HAP lateral distance (km)
HAP payload antenna: WiMAX WiMAX
G = 7 dBi antenna 1 antenna 2
G = 13 dBi Optical
window
Figure 5: Derived UL modulation level (“rate”) versus HAP
displacement for G = 15 dBi ground antenna. Figure 6: Schematic of payload physical layout.

than the downlink (BS output power). Should the UL fail,

then the system becomes inoperative. Since one CPE acts
as the backhaul, it is particularly important to maintain a
reasonable data rate on that link, since this acts effectively
as a bottleneck on the whole WiMAX HAP network. For
present purposes with only two or three CPEs, this is not a
major problem, but for a future or commercial deployment
a different architecture, with dedicated backhaul spectrum,
would be preferable.

6. PAYLOAD Figure 7: Payload interior components.

The combined WiMAX and optical payloads are to be

housed in a spherical pressure chamber for deployment
in the stratosphere. The sphere diameter is 750 mm. This to service the gimballed camera and up to a further three
layout dictates much of the mechanical design, which is fixed (and expendable) external video cameras mounted
illustrated in Figure 6, where the important payload com- outside the pressure chamber. After compression the video is
ponents are mounted above and below a central circular streamed over the WiMAX network to ground based clients.
aluminium frame which comprises numerous access holes The optical subpayload is housed low in the payload vessel
to facilitate assembly. The optical subpayload comprises a and behind an optical window. The remaining payload skin
gimbal mounted camera which has integral stabilisation and is optically opaque. The WiMAX antennas are mounted
control system. This is a “TASE” [11] unit procured from externally to the vessel and connected via hermetically sealed
Cloud Cap Technology Inc., Hood River, USA who has microwave bulkhead connectors. The equipment is cooled by
developed the unit for unmanned aerial vehicles (UAVs). It recirculating the air in the chamber at atmospheric pressure,
features a 112 mm diameter gimbal turret, 0.05◦ pointing thus emulating a pressure and temperature environment very
resolution in a package weighing 0.9 kg. The bespoke server similar to that at ground level.
control unit houses the failsafe microcontroller, power The components of the “upper” payload are shown in
management and conditioning, environment monitoring, Figure 7 where the base station and server control units are
network infrastructure and two pc servers. PC server 1 visible above the circular aluminium frame, below which are
running a Linux OS provides locally hosted web, email, and the WiMAX RF unit and the optical subpayload. The darker
ftp services, which allows the demonstration of these without hemispheres of the pressure vessel are also discernible.
the requirement for a high speed backhaul connection to A view of the lower side of the payload is shown in
the internet. Server 1 also hosts web proxy services to Figure 8 where the optical window is apparent, although the
demonstrate their ability to reduce user download times exterior WiMAX antennas are not yet attached. The payload
and to reduce the backhaul traffic. PC server 2, running maximum power consumption is 180 W, including the DC
embedded Microsoft XP OS, hosts video acquisition software power converters.
8 EURASIP Journal on Wireless Communications and Networking

Figure 8: Exterior of lower payload.

Figure 10: Payload components for first airborne trial.

8. GROUND-BASED USER TRIAL

Following the successful airborne trials the WiMAX and
optical payloads were fully integrated into the spherical
pressure chamber and, complete with ancillary equipment
such as power supplies and UHF radio modem, prepared
for a further, longer duration trial. During August 2007 the
payload package was transported and temporarily installed
within a north-east facing hotel room at the Pilatus Moun-
tain complex, Switzerland [12]. From this vantage point at
2132 m (7000 ft) altitude the two small circularly polarised
Figure 9: Helicopter-mounted payload. antennas mounted vertically on the outside of the building
window had a clear line of sight path to the ground station
located in Hünenberg, 24.4 km distant. The ground station
consisted of two CPEs, one acting as the gateway to the wired
internet and one as the “remote terminal” as in Figure 1. The
7. FIRST AIRBORNE TRIALS trial consisted of over 30 hours of continuous use with the
remote CPE providing “real” users with access to the internet
The payload components were prepared for installation via wired and WiFi connectivity. Users had access to live
on a helicopter (see Figure 9) for a first airborne trial video streams provided via the onboard cameras as well as
in Switzerland during July 2007. Being a nonstratospheric a host of locally (to the payload) stored media. During the
trial, the pressure vessel was not used at this stage. The trial the payload was remotely managed by staff via the UHF
two WiMAX antennas were mounted so as to provide a telecommand and control link demonstrating the ability to
downward beam and a rear facing beam since the helicopter enter very low power states and to reactivate the payload
would fly to a 40 km ground distance but not exceed an systems.
altitude of about 2 km: these are visible, respectively, as the
black circular and square antennas in Figure 10, and below 9. CONCLUSIONS
the former, on the left-hand side, the very compact “TASE”
camera turret is clearly seen. The development of a WiMAX payload for a stratospheric
The helicopter flew two missions whose flight plans HAP trial has been reported, where commercially available
are illustrated in Figure 11. The WiMAX link was kept equipment has been used as far as possible and integrated
operational using the wider beamwidth, downward facing with bespoke equipment. We used Alvarion’s Breezemax base
antenna during each takeoff and climb phase, and the station and associated customer equipment operating in
narrower beamwidth rear facing antenna during the outward bands close to 3.5 GHz. The uplink and downlink data rates
flights. The link was thus unavailable during the return are asymmetric due to an 8 dB difference in RF unit output
flights. power (resp., 20 dBm and 28 dBm). In terrestrial evaluation
The performance of the uplink during these missions is and using standard linear polarised antennas of aggregate
summarised in Figure 12 by plotting the logged modulation gain 32 dBi, we obtained data throughputs of 8 Mbps and
level. This frequently peaked at rate 8 at distances up to 9 Mbps, respectively, for UL and DL at a distance of 20 km:
28 km. The “bursty” nature of the logged uplink rate is these are in fact the highest rates we have observed from the
typical and also seen in Table 3. The sudden fluctuations equipment. HAP displacement and pitch/roll considerations
in rate seen during the flights, as shown in Figure 12, are lead to an optimum HAP payload antenna directivity of
due to the combined effects of the bursty UL demand and 16 dBi and ground antenna directivity of 18 dBi. Assuming
link outages during helicopter manoeuvring; the latter effect 50% antenna efficiency, this gives an aggregate antenna gain
being a consequence of the relatively narrow antenna beams. of 28 dBi: a link budget calculation predicts that the UL
John Thornton et al. 9

remote payload control during a 30-hour period over a 24.4-

Ground station
Zug km distance.
These trials have demonstrated the potential for WiMAX
services from aerial platforms, and validated design aspects
of the HAP communications payload. Subsequent phases of
Lucerne the trials are planned with the StratXX airship operating at
around 2-km altitude.

ACKNOWLEDGMENTS
The authors thank Steve Jones and the nonoperational
license team at Ofcom in the UK. They also thank Technical
Central Switzerland
Support Services at Department of Electronics, University of
20 km
Altdorf York.

WiMAX: REFERENCES
Flight 1 Flight 2
[1] N. J. Colella, J. N. Martin, and I. F. Akyildiz, “The HALO
Operational Operational
networkTM ,” IEEE Communications Magazine, vol. 38, no. 6,
Non-operational Non-operational
pp. 142–148, 2000.
Figure 11: Helicopter flight plan. [2] T. C. Tozer and D. Grace, “High-altitude platforms for wireless
communications,” Electronics and Communication Engineering
Journal, vol. 13, no. 3, pp. 127–137, 2001.
[3] J. Thornton, D. Grace, C. Spillard, T. Konefal, and T. C.
Tozer, “Broadband communications from a high-altitude
Flight 1 Flight 2 platform: the European HeliNet programme,” Electronics and
40 8
Altitude (1000 ft)/distance (km)

Communication Engineering Journal, vol. 13, no. 3, pp. 138–

35 7 144, 2001.
30 6 [4] R. Miura and M. Suzuki, “Preliminary flight test program
Modulation rate

25 5 on telecom and broadcasting using high altitude platform

stations,” Wireless Personal Communications, vol. 24, no. 2, pp.
20 4
341–361, 2003.
15 3 [5] www.elec.york.ac.uk/comms/projects/helinet/index.html.
10 2 [6] http://www.capanina.org.
6 1 [7] M. Mohorcic, D. Grace, G. Kandus, and T. C. Tozer, “Broad-
4
2 band communications from aerial platform networks—an
0 10 20 30 40 50 60 70 80 90 overview of CAPANINA,” in Proceedings of the 13th IST Mobile
Mission time (minutes) and Wireless Communications Summit, pp. 257–261, Lyon,
France, June 2004.
Altitude [8] D. Grace, M. H. Capstick, M. Mohorcic, J. Horwath, M. B.
UL mod. rate Pallavicini, and M. Fitch, “Integrating users into the wider
Distance broadband network via high altitude platforms,” IEEE Wireless
Figure 12: Helicopter flights: uplink rate, altitude and link distance Communications, vol. 12, no. 5, pp. 98–105, 2005.
versus time. [9] M. Bobbio Pallavicini, J. Horwath, M. Capstick, et al.,
“Test Results Summary Report, FP6 CAPANINA Project,”
Doc Ref: CAP-22a-WP44-CGS-PUB-01, September 2006,
http://www.capanina.org/documents/CAP-D22a-WP44-CGS
-PUB-01.pdf.
should run at 16QAM modulation and 3/4 code rate offering [10] J. Thornton, D. Grace, M. H. Capstick, and T. C. Tozer,
8 Mbps net in each 3.5 MHz bandwidth channel for an HAP “Optimizing an array of antennas for cellular coverage from
at height 20 km within a 5-km position sphere. a high altitude platform,” IEEE Transactions on Wireless
A payload comprising WiMAX base station, microcon- Communications, vol. 2, no. 3, pp. 484–492, 2003.
troller, PC servers and application software, camera, and [11] http://www.cloudcaptech.com/gimbal tase.shtm.
gimbal subpayload, RF unit, RF switch and antennas, was [12] http://www.pilatus.ch.
constructed at the University of York for integration into
a spherical pressure chamber for stratospheric trials to be
carried out later in 2007 or in 2008. The payload was
evaluated in an airborne trial in Switzerland in July 2007
where UL and DL WiMAX links performed as expected at
distances up to 36 km during helicopter flights. A longer
duration terrestrial trial of the integrated payload followed
in August 2007 demonstrating successful link operation and
 Photographȱ©ȱTurismeȱdeȱBarcelonaȱ/ȱJ.ȱTrullàs

Preliminaryȱcallȱforȱpapers OrganizingȱCommittee
HonoraryȱChair
The 2011 European Signal Processing Conference (EUSIPCOȬ2011) is the MiguelȱA.ȱLagunasȱ(CTTC)
nineteenth in a series of conferences promoted by the European Association for GeneralȱChair
Signal Processing (EURASIP, www.eurasip.org). This year edition will take place AnaȱI.ȱPérezȬNeiraȱ(UPC)
in Barcelona, capital city of Catalonia (Spain), and will be jointly organized by the GeneralȱViceȬChair
Centre Tecnològic de Telecomunicacions de Catalunya (CTTC) and the CarlesȱAntónȬHaroȱ(CTTC)
Universitat Politècnica de Catalunya (UPC). TechnicalȱProgramȱChair
XavierȱMestreȱ(CTTC)
EUSIPCOȬ2011 will focus on key aspects of signal processing theory and
TechnicalȱProgramȱCo
Technical Program CoȬChairs
Chairs
applications
li ti as listed
li t d below.
b l A
Acceptance
t off submissions
b i i will
ill be
b based
b d on quality,
lit JavierȱHernandoȱ(UPC)
relevance and originality. Accepted papers will be published in the EUSIPCO MontserratȱPardàsȱ(UPC)
proceedings and presented during the conference. Paper submissions, proposals PlenaryȱTalks
for tutorials and proposals for special sessions are invited in, but not limited to, FerranȱMarquésȱ(UPC)
the following areas of interest. YoninaȱEldarȱ(Technion)
SpecialȱSessions
IgnacioȱSantamaríaȱ(Unversidadȱ
Areas of Interest deȱCantabria)
MatsȱBengtssonȱ(KTH)
• Audio and electroȬacoustics.
• Design, implementation, and applications of signal processing systems. Finances
MontserratȱNájarȱ(UPC)
Montserrat Nájar (UPC)
• Multimedia
l d signall processing andd coding.
d
Tutorials
• Image and multidimensional signal processing. DanielȱP.ȱPalomarȱ
• Signal detection and estimation. (HongȱKongȱUST)
• Sensor array and multiȬchannel signal processing. BeatriceȱPesquetȬPopescuȱ(ENST)
• Sensor fusion in networked systems. Publicityȱ
• Signal processing for communications. StephanȱPfletschingerȱ(CTTC)
MònicaȱNavarroȱ(CTTC)
• Medical imaging and image analysis.
Publications
• NonȬstationary, nonȬlinear and nonȬGaussian signal processing. AntonioȱPascualȱ(UPC)
CarlesȱFernándezȱ(CTTC)
Submissions IIndustrialȱLiaisonȱ&ȱExhibits
d i l Li i & E hibi
AngelikiȱAlexiouȱȱ
Procedures to submit a paper and proposals for special sessions and tutorials will (UniversityȱofȱPiraeus)
be detailed at www.eusipco2011.org. Submitted papers must be cameraȬready, no AlbertȱSitjàȱ(CTTC)
more than 5 pages long, and conforming to the standard specified on the InternationalȱLiaison
EUSIPCO 2011 web site. First authors who are registered students can participate JuȱLiuȱ(ShandongȱUniversityȬChina)
in the best student paper competition. JinhongȱYuanȱ(UNSWȬAustralia)
TamasȱSziranyiȱ(SZTAKIȱȬHungary)
RichȱSternȱ(CMUȬUSA)
ImportantȱDeadlines: RicardoȱL.ȱdeȱQueirozȱȱ(UNBȬBrazil)

P
Proposalsȱforȱspecialȱsessionsȱ
l f i l i 15 D 2010
15ȱDecȱ2010
Proposalsȱforȱtutorials 18ȱFeb 2011
Electronicȱsubmissionȱofȱfullȱpapers 21ȱFeb 2011
Notificationȱofȱacceptance 23ȱMay 2011
SubmissionȱofȱcameraȬreadyȱpapers 6ȱJun 2011

Webpage:ȱwww.eusipco2011.org

View publication stats