Académique Documents
Professionnel Documents
Culture Documents
Galileo Lite;
GLONASS
Tuned; GPS L5
Set to Go
THE BUSINESS
Security for
Military,
Mass Market
EXPERT ADVICE
IMES Evaluated
SYSTEM DESIGN
SBAS Interferes
BUSINESS OUTLOOK
VOLUME 20 | NUMBER 4
SOLUTIONS AND PRODUCTS FOR A
NEW ERA IN SATELLITE NAVIGATION
... improved GLONASS ...
𰀩𰁎𰁃𰁒𰁅𰁁𰁓𰁅𰀀𰁙𰁏𰁕𰁒𰀀𰁐𰁒𰁏𰁄𰁕𰁃𰁔𰁉𰁖𰁉𰁔𰁙𰀀𰁗𰁉𰁔𰁈𰀀 𰀳𰁁𰁔𰁉𰁓𰁆𰁙𰀀𰁙𰁏𰁕𰁒𰀀𰁒𰁅𰁓𰁅𰁁𰁒𰁃𰁈𰀀𰁐𰁒𰁏𰁊𰁅𰁃𰁔𰀀𰁎𰁅𰁅𰁄𰁓𰀀𰁗𰁉𰁔𰁈𰀀
𰀧𰀰𰀳𰀀𰀬𰀑𰀌𰀀𰀬𰀒𰀌𰀀𰀬𰀕𰀀𰁃𰁏𰁍𰁍𰁅𰁒𰁃𰁉𰁁𰁌𰀀𰁐𰁒𰁏𰁄𰁕𰁃𰁔𰁓𰀁 𰁃𰁕𰁓𰁔𰁏𰁍𰁉𰁓𰁁𰁂𰁌𰁅𰀀𰀧𰁁𰁌𰁉𰁌𰁅𰁏𰀀𰀥𰀑𰀌𰀀𰀥𰀕𰀌𰀀𰀥𰀖𰀀𰁓𰁏𰁌𰁕𰁔𰁉𰁏𰁎𰁓𰀁
NavX®-NCS – RF Navigation
Constellation Simulator
Now Supports Galileo CBOC
and Japanese QZSS Signals
𰀴𰁈𰁅𰀀𰁬𰁒𰁓𰁔𰀀𰀧𰀰𰀳𰀏𰀧𰁁𰁌𰁉𰁌𰁅𰁏𰀀𰀲𰀦𰀀𰁃𰁏𰁎𰁓𰁔𰁅𰁌𰁌𰁁𰁔𰁉𰁏𰁎𰀀𰁓𰁉𰁍𰁕𰁌𰁁𰁔𰁏𰁒𰀌𰀀
𰁓𰁕𰁐𰁐𰁏𰁒𰁔𰁉𰁎𰁇𰀀𰁁𰁌𰁌𰀀𰁆𰁒𰁅𰁑𰁕𰁅𰁎𰁃𰁉𰁅𰁓𰀀𰁉𰁎𰀀𰁏𰁎𰁅𰀀𰁂𰁏𰁘𰀌𰀀𰁒𰁅𰁄𰁅𰁬𰁎𰁅𰁓𰀀𰁔𰁈𰁅𰀀
𰁆𰁒𰁏𰁎𰁔𰁉𰁅𰁒𰀀𰁏𰁆𰀀𰀲𰀦𰀀𰁃𰁏𰁎𰁓𰁔𰁅𰁌𰁌𰁁𰁔𰁉𰁏𰁎𰀀𰁓𰁉𰁍𰁕𰁌𰁁𰁔𰁏𰁒𰀀𰁔𰁅𰁃𰁈𰁎𰁏𰁌𰁏𰁇𰁙𰀎
𰁳𰀀𰀀𰀧𰀰𰀳𰀀𰀬𰀑𰀌𰀀𰀬𰀒𰀌𰀀𰀬𰀕
𰁳𰀀𰀀𰀧𰀡𰀬𰀩𰀬𰀥𰀯𰀀𰀥𰀑𰀌𰀀𰀥𰀕𰀌𰀀𰀥𰀖
𰁳𰀀𰀀𰀧𰀬𰀯𰀮𰀡𰀳𰀳𰀀𰀬𰀑
Test Solutions
𰁳𰀀 𰀱𰀺𰀳𰀳𰀀𰀬𰀑
Receiver 𰁳𰀀𰀀𰀵𰁐𰀀𰁔𰁏𰀀𰀑𰀐𰀘𰀀𰁃𰁈𰁁𰁎𰁎𰁅𰁌𰁓𰀀𰁉𰁎𰀀𰁔𰁈𰁅𰀀𰁐𰁒𰁏𰁆𰁅𰁓𰁓𰁉𰁏𰁎𰁁𰁌𰀀𰁖𰁅𰁒𰁓𰁉𰁏𰁎
𰁳𰀀𰀀𰀵𰁐𰁇𰁒𰁁𰁄𰁅𰀀𰁃𰁁𰁐𰁁𰁂𰁉𰁌𰁉𰁔𰁙𰀀𰁂𰁙𰀀𰁁𰁄𰁄𰁉𰁎𰁇𰀀𰁃𰁈𰁁𰁎𰁎𰁅𰁌𰀀𰁍𰁏𰁄𰁕𰁌𰁅𰁓
Software Tools 𰁳𰀀𰀀𰀀𰀦𰁒𰁅𰁅𰀀𰁓𰁅𰁌𰁅𰁃𰁔𰁉𰁏𰁎𰀀𰁏𰁆𰀀𰁆𰁒𰁅𰁑𰁕𰁅𰁎𰁃𰁙𰀀𰁁𰁎𰁄𰀀𰁓𰁙𰁓𰁔𰁅𰁍
www.ifen.com
sales@ifen.com
APRIL 2009
VOL. 20, NUMBER 4
gpsworld.com
THE BUSINESS 14
Security Paramount, Strengthened
by Location; Broadcom, SiRF Top
ABI’s GPS Chip Vendor Ranking;
ITT Awards Symmetricom Timing
Contract for GPS III; more
BUSINESS
OUTLOOK 27
How the Stimulus Plan
Benefits Our Industry
How will the American Recovery
and Reinvestment Act of 2009
affect our industry?
By Stephen Colwell
SBAS Interference 37
OPINIONS & Worst-Case Scenario
Some applications using data collected by fixed- or low-dynamic
DEPARTMENTS WAAS or other SBAS receivers may encounter a tracking error due
to C/A code cross-correlation.
By L. Lestarquit, J-L. Issler, O. Nouvel, M. Sihrener, G. Lamontagne,
Out in Front 4 J.C. Guay, R. Landry, O. Julien, C. Macabiau, and M. Nouvel-Malicorne
Talking Heads
By Alan Cameron
Expert Advice 6
Indoor Message System Evaluated
Precise Point Positioning 43
By Börje Forssell
A Powerful Technique with a Promising Future
The IMES indoor messaging
Is there a viable alternative to real-time kinematic (RTK) positioning?
system aims to provide seamless
In this month’s column, we take a look at the technique of precise
positioning at the room level.
point positioning (PPP). How close is PPP to prime time?
By Sunil Bisnath and Yang Gao
Marketplace Classifieds 49
W
e now feel ourselves dragged much akin to that relic of bygone
Editor in Chief Alan Cameron | acameron@questex.com
into the second decade of decades, the cocktail party. Managing Editor Tracy Cozzens | tcozzens@questex.com
the 21st century — some of Talk. Second method, I happened Art Director Annie Leung | aleung@questex.com
us kicking and screaming. An edict has across a fellow I know, who coinciden- EDITORIAL OFFICES
arrived from on high: explore the use of tally authored two cover stories in this 201 Sandpointe Avenue, Suite 500,
Santa Ana, CA 92707-8716 USA
new media to further our mission and magazine and is otherwise active in the 714-338-6700 | Fax 714-338-6717
increase our numbers, whether they GNSS community. Do you find social www.gpsworld.com | info@gpsworld.com
count revenue, pages, or web traffic. media any use in your professional life, CONTRIBUTING EDITORS
Pick a number. Any number. Now industry, community? I asked. Innovation Richard Langley | lang@unb.ca
LBS Kevin Dennehy | kdennehy@questex.com
increase it. “I generally don’t,” he wrote back, Mass Market OEM Stephen Colwell | scolwell@questex.com
Well and good. A former marketing “have any expectations for Facebook as a Military/Government Don Jewell | djewell@questex.com
Precision OEM Rob Lorimer | rlorimer@questex.com
communications account exec, I’m all professional tool. Some ‘grown-up’ Survey Eric Gakstatter | egakstatter@questex.com
for exploring the possible benefits of new versions of Facebook such as Linked-In Transport/Avionics Bill Thompson | bthompson@questex.com
Utilities/Communications Janice Partyka | jpartyka@questex.com
media, and the new ways of thinking and Plaxo are going for that market. Even
ADVERTISING
that they may instill. Exploring, talking those, I’ve been building a professional
Publisher Kristina Panter
with those who have gone before, and network, but find it of little real use. If I kpanter@questex.com | 714-338-6733
then evaluating the overall findings. want to e-mail somebody, I e-mail them. Eastern Region/International Account Manager Michael Eager
What follows does not constitute I don’t need a web-based thing for that. meager@questex.com | 714-338-6738
only and mere navel-gazing. Replace my “My corporate communications Western Regional Account Manager Jeannie Hortness
jhortness@questex.com | 714-338-6732
enterprise with yours as you read. Would policy limits the amount I can just blog
Classified Account Executive Candi Eager
your company or organization, or the about work-related stuff. I couldn’t post ceager@questex.com | 714-338-6708
pictures or videos about my work if I Marketing Coordinator Sarah Joy Obaña
Would you get a reasonable ROI wanted to. sobana@questex.com | 714-338-6763
“Finally, I do this Facebook thing to Sales Administrator Marisa Strange
of your time in Twitter, Linked-In, get away from the geek/nerd community mstrange@questex.com | 714-338-6766
ADVERTISING OFFICES
that I’m immersed in all day. If I needed 201 Sandpointe Avenue, Suite 500, Santa Ana, CA 92707-8716 USA
IM, Facebook, YouTube? more outlets for that, there are plenty of 714-338-6700 | Fax 714-338-6717
web resources that would be better. PUBLISHING SERVICES
GNSS industry as a whole, get a reason- “[Another GPSW author] related Production Manager Anna Abbott | aabbott@questex.com
Audience Development Manager Deborah Gullian |
able return on investment of time spent an interesting observation when I dgullian@questex.com
in Twitter, Linked-In, instant messaging first started using Facebook. To him PRODUCTION OFFICE 306 West Michigan St., Suite 200,
(IM) and other chat-forms, Facebook, it was disturbing when professional Duluth, MN 55802, USA; 218-279-8840
MySpace, YouTube? acquaintances started asking to be CIRCULATION/SUBSCRIBER SERVICES
Explore. My admittedly minimal friends on Facebook because it was like gpsworld@halldata.com | USA: 866-344-1315 (toll free)
International: 1-847-763-9594
casting about in Linked-In and Facebook ‘worlds colliding’.” QUESTEX MEDIA GROUP, INC.
suggests that Linked-In suits admirably Evaluate. I’ll keep an open mind President & CEO: Kerry C. Gumas
for job-finding and consultancy-fishing, but not invest more time at present. EVP & CFO: Tom Caridi
not meaningful lead- or traffic-genera- Two years ago, I extolled the virtues EVP: Tony D’Avino
VP, Human Resources: Diane M. Evans
tion. There isn’t enough room or context of blogging to deepen contact with
QUESTEX WORLDWIDE HEADQUARTERS
for content, because it was designed for customers and other influentials in your 275 Grove Street, Newton, MA 02466, USA
the former two goals, not the latter two. sphere (“Dagbloggit,” February 2007). 617-219-8300 | Toll-Free 888-552-4346 | Fax 617-219-8310
Facebook — you can go by its name. I stand by that; but so far, it doesn’t
MANUSCRIPTS: GPS World welcomes unsolicited articles but cannot be held responsible
Mostly about faces, and the superfici- extend to FB. for their safekeeping or return. Send to: 201 Sandpointe Avenue, Suite 500, Santa Ana,
CA 92707-8716 USA. Every precaution is taken to ensure accuracy, but publishers cannot
ality that accompanies appearances. New Media. For now, I’m going with accept responsibility for the accuracy of information supplied herein or for any opinion
expressed. REPRINTS: Reprints of all articles are available (500 minimum). Contact
Can be fun, pass the time, keep up or a new medium that’s exactly about as old 800-290-5460, ext. 100, e-mail GPSWorld@reprintbuyer.com. DIRECT MAIL LIST
RENTAL: Ilene Schwartz, Kroll Direct, 216-371-1667, fax 216-371-1669 e-mail Ilene@
attempt to strike up with acquaintances as I am. Check out our website for videos. krolldirect.com. SUBSCRIBER SERVICES: To subscribe, change your address, and all
old and new. But not every human Not Youtube, exactly. Short interviews other services, e-mail gpsworld@halldata.com or call 866-344-1315 (1-847-763-9594 outside
the U.S.). PERMISSIONS: Contact 800-494-9051 ext. 100 or questexpermissions@
pursuit naturally, eventually, or auto- with leaders of the GNSS community. reprintbuyer.com. INTERNATIONAL LICENSING: Contact e-mail info@gpsworld.com.
ACCOUNTING OFFICE and OFFICE OF PUBLICATION: 306 West Michigan
matically becomes a business pursuit Starting with Dr. Brad Parkinson and St., Ste 200, Duluth, MN 55802, USA.
GPS WORLD does not verify any claims or other information appearing in any of the
if practiced long enough. Facebook is some key Galileo figures. View on. advertisements contained in the publication and cannot take any responsibility for any losses
or other damages incurred by readers in reliance on such content.
Published monthly
Letters to Editor invited: email to eic@gpsworld.com.
www.whiteedc.com
P
ublic and commercial services are being tried, but with no definitive position of the center of its cell coverage
such as search-and-rescue, solution in sight. zone, or linking the receiver to a database
firefighting, and location- corresponding to an identifier. Instead
based services (LBS) require accurate The IMES Concept of the ephemeris data, clock corrections,
indoor position determination. Users The Japan Aerospace Exploration ionospheric parameters, and so on
face the fundamental problem of Agency, GNSS Technologies, and contained in the GPS message, the
knowing location at the room level, Lighthouse Technology and Consulting IMES navigation message periodically
requiring accuracies on the order of a have proposed and developed an dispatches position broadcasts and
few meters. Because of the need for floor indoor messaging system (IMES) to additional information in a similar
determination, vertical accuracy can be solve the availability issue. Composed format.
more important than horizontal. of transmitters, GPS receivers with The concept for seamless transition
modified firmware embedded in cell between indoor and outdoor use is built
phones, and systems of servers, IMES on three further requirements:
aims to provide seamless positioning 䡲 low power consumption,
anywhere in a covered area. Hitachi 䡲 low transmitter cost, and
Ltd., Systems Development Laboratory 䡲 no harmful interference to other
is also working on such a system, as users of GPS or QZSS.
shown in FIGURE 1. IMES assumes that positioning
IMES uses satellite signals outdoors accuracies on the order of 10 meters
in the usual way, while using signals will satisfy users, who just want to know
from IMES transmitters indoors, where where they are in rooms of moderate
For indoor applications, stable and satellite signal quality is strongly reduced. size, shopping areas, parking garages,
IMES signal structure is similar to that and so on. Stable and reliable position
reliable position information means of GPS satellite signals, except for the information means more than high
contents of the navigation message. Thus, accuracy.
more than high accuracy. the same receiver can be used for both.
An IMES transmitter sends an RF Signals and Codes
signal similar to that of GPS and the IMES uses the same L1 center frequency
The United States and Japan have Japanese Quasi-Zenith Satellite System as GPS and QZSS, and the same BPSK
passed emergenc y-ca ll legislation (QZSS), giving its 3-D position, the modulation. IMES dedicated spread-
requiring third-generation cell phones
to have functions determining their
positions. Rapidly increasing numbers of
cell phones with built-in GPS capabilities
stimulate the LBS market, which is
expected to be the dominating market
for GNSS receivers in the future.
Si g n i f ic a nt proble m s re m a i n,
however. Standard GPS receivers do
not work well indoors, giving position
errors up to several tens of meters — if
a position fix is possible at all. Assisted
GPS (AGPS) improves the situation, but
makes equipment more complicated and
expensive, while reliability and accuracy
may still not suffice. Local solutions
based on RFID, Wi-Fi, ultra-wideband,
pseudolites, and others have been and Å FIGURE 1 IMES principles, from a Hitachi presentation at Tokyo GNSS Symposium
6 GPS World | April 2009 www.gpsworld.com
EXPERT ADVICE
spectrum codes are from the same outdoors and would not create harmful receiver imply a signal-to-interference
family of Gold codes as GPS and QZSS interference in satellite receivers. ratio up to a maximum of ⫺11 to ⫺14 dB
(numbers 173–182 from the C/A-code However, the IMES concept foresees (referred to the autocorrelation peak),
assignment table). It is assumed that the possibility of having transmitters in depending on the transmitted power
signals similar to L1C can be used in many rooms in large buildings, shopping and the attenuation situation. Obviously,
the future. These dedicated codes are areas, and so on, which may imply that in some situations, ordinary reception,
re-used repeatedly but without using the ordinary GPS receivers outdoors in and above all, acquisition of GPS L1
same code as neighboring transmitters. such an area receive emissions from a C/A-code signals can be blocked or at
The IMES receiver uses the codes only considerable number of transmitters. least severely hampered. In addition, a
for de-spreading the spread-spectrum portable IMES transmitter would make
modulation and as a step to decoding the The Interference Problem a good short-range GPS jammer. Using
navigation message, including transmitter It is of crucial importance that IMES new GPS signals such as L2C with better
identification. No pseudorange or time signals do not reduce performance for cross-correlation margins would reduce
determination is necessary, because the ordinary GPS receivers used outside the the problem.
desired position is read directly out of the intended IMES coverage area. Because
navigation message. This gives a much IMES uses the same L1 frequency and Conclusions
simpler equipment architecture than the same type of Gold codes as GPS IMES represents a new idea in the field
when using pseudolites. satellites, the interference will be of of position determination in confined
the usual Gold-code cross-correlation areas. However, the advantages should
Transmitter Power type. If the received power from two not be gained at costs of ordinary
The transmitted power varies depend- different transmitters is the same, the GPS/QZSS users. For this reason,
ing on the location and the distance to cross correlation levels of Gold codes transmitted power levels should be set
neighboring transmitters. Japanese radio are between ⫺21.8 and ⫺23.6 dB lower than those indicated here and
regulations for license-freesignals require (depending on Doppler difference) as proposed by the system developers,
that the signal level at 3 meters from the below the autocorrelation peak. to reduce the probability of harmful
transmitter stay below 35 µV/meter, The actual level of cross-correlation interference. Because of the increasing
limiting the radiated power density from interference in a GPS receiver operating availability of high-sensitivity receivers,
IMES transmitters. in an IMES environment is determined it is possible to obtain the wanted
The radiated power level determines by the power received from each IMES performance with lower power levels.
the coverage zone of a particular IMES transmitter in question and the number The IMES group seems to be aware
transmitter: larger coverage implies more of these transmitters. of this, conducting various experiments
power. Because of the requirement for IMES signals are attenuated by walls, and analyses for optimum setting
compatibility with outdoor GPS signals, floors, roofs, and other obstacles between and design. We have to wait for final
IMES power levels received indoors are the transmitter and the GPS receiver. results, but received power levels around
expected to be found in the interval Such attenuation varies between just a ⫺140 dBm (10 dB below GPS satellite
between ⫺126 dBm and ⫺130 dBm few dB (window or open door) to several signals) might still do the job. In all cases,
at a distance of about 5 meters from the tens of dB (floors/ceilings of reinforced IMES transmission levels must be
transmitter antenna. c oncrete ) . Me a su rement s i n ou r adapted to the local situation. We must
The inventors of IMES have conducted university building in Trondheim have keep in mind, however, that IMES
experiments with these power limits in given attenuation values such as 20 dB reception itself is vulnerable to external
mind, using transmitter signal levels through a gypsum wall, 35 dB through a interference (including signals from GPS
from ⫺64 to ⫺76 dBm. The former is ceiling of reinforced concrete, and 66 dB satellites), so transmitted power cannot
the maximum level allowed for license- through four such ceilings. be arbitrarily low.
free signals. The working range of IMES Each IMES signal received increases
is assumed to be 2–10 meters from the the cross-correlation interference level in BÖRJE FORSSELL is professor of navigation in
the Department of Telecommunications at
transmitting antenna, and ⫺64 dBm at the GPS receiver. From an interference the Norwegian University of Science and
the transmitter means about ⫺120 dBm point of view, these signals should be Technology in Trondheim, Norway.
at the receiving antenna at a distance regarded as noise, and the resulting total
For reference documents, see the online
of 10 meters. Accordingly, ⫺76 dBm variance of these noise-like signals is the version of this article at www.gpsworld.com/imes.
transmitted means about ⫺132 dBm at sum of the variances of all the interfering
the receiving antenna. signals, that is, the sum of the cross-
This latter level is close to what can correlations. Thus, 10 IMES transmitters
be expected for a GPS satellite signal in the immediate neighborhood of a GPS
8 GPS World | April 2009 www.gpsworld.com
SYSTEM
THE
contract by the end of the month. their success, and DAGR will continue
DAGR Extended This will provide for an additional to provide real-time PNT services to
51,200 selective availability anti- the fingertips of the soldiers fighting
From a Space & Missile Systems spoofing module-based DAGRs the ongoing war on terror.’
Center media release, March 17: and associated accessories. This “Under the current contract,
“The GPS Wing awarded a represents a $200 savings per unit Rockwell Collins has delivered more
sole-source follow-on production over the previous contract. than 290,000 DAGRs to the armed
contract for Defense Advanced “‘I am extremely pleased with the forces of the United States and
GPS Receiver with a ceiling of outstanding accomplishments of 22 Foreign Military Sales customers.
$450 million to Rockwell Collins Inc. both parties to negotiate a win-win DAGR is the current state-of-the-art
of Cedar Rapids, Iowa, March 10. agreement,’ said Col. Dave Madden, military handheld GPS receiver ...
“The first production option commander, GPS Wing. ‘The The DoD has recognized the
buy of $87 million will be put on warfighter will directly benefit from (Continued on page 12)
10 GPS World | April 2009 www.gpsworld.com
THE SYSTEM
Moerdijk, Oostflakkee or wherever; you can now map cities as if you’re right there. Thanks to
Cyclomedia’s 360-degree panoramic digital mapping with NovAtel precise positioning technology
onboard, the world appears in 3-D because, well, it is. Incorporating NovAtel’s SPAN GPS+Inertial
Measurement Unit system, accurate and robust positioning data is possible even in challenging urban
conditions where GPS alone is limited.
3-D mapping is just one of many worldwide applications successfully utilizing NovAtel’s precise
positioning products.
1-800-NOVATEL (U.S. & Canada) or 403-295-4900 | Europe +44 (0) 1993 852-436 | SE Asia & Australia +61 (2) 8668 4073 | sales@novatel.com
» TRANSPORTATION
» PROFESSIONAL OEM
V
ehicle-to-vehicle (V2V) communication-based ad- (ACC) systems, side blind-zone alert systems, and many more.
vanced driver assistance systems (ADAS) enable or Most of these applications require specific knowledge about
enhance a suite of safety features by combining GPS vehicles in close proximity, typically their speed, acceleration,
and short-range wireless communication. The allocation of and heading.
the 5.9 MHz radio spectrum for highway safety and traffic In an ACC implementation, radar, lidar, or other technologies
efficiency applications using Dedicated Short Range Com- track target vehicles in the field of view. However, this requires
munications (DSRC) has given automotive manufacturers the multiple sensors to cover different regions around the vehicle and
ability to go beyond traditional sensor-based safety features to to address the range requirements of different applications. For
provide drivers with another sense. DSRC enables vehicles to instance, forward- or backward-looking safety applications typi-
talk to surrounding vehicles and maintain a comprehensive cally require the pairing of short- and long-range radar sensors
situational awareness of roadway traffic. and separate sensors to provide coverage of side blind zones as il-
General Motors first publicly demonstrated a fleet of com- lustrated in FIGURE 1 (left panel), showing tracked and untracked
municating vehicles with driver warnings and automatic vehicle vehicles using dark and light shades respectively.
braking in 2005. Recently, General Motors has further developed These traditional sensors do not provide coverage beyond im-
the system with its OnStar and StabiliTrak technologies. The mediate line-of-sight, which may limit the effectiveness of these
latest GM V2V fleet is fully integrated with production tech- applications. For instance, in a hard-brake event that occurs far
nologies with the exception of DSRC. Its built-in communica- ahead, following drivers may not see the event until the vehicle
tion system can also interface with roadway infrastructure and directly ahead takes a responsive action. V2V systems can provide
support a range of new vehicle-to-infrastructure (V2I) features instantaneous communication between vehicles, even in situa-
as they become available. tions where an event is obscured by a sharp turn, a large vehicle,
An accurate, reliable, and affordable positioning system, flex- or some other obstruction.
ible enough to evolve with emerging V2V and V2I application If vehicles carry location awareness technology such as GPS
requirements, obviously constitutes a prerequisite for these sys- and the ability to talk to each other, all vehicles and even the
tems to work. This system is based on an efficient over-the-air infrastructure elements such as traffic-control devices could
(OTA) messaging scheme and an onboard system design that can communicate and incorporate situation awareness onboard. In
be scaled to fit the needs of a range of in-vehicle safety features. Figure 1’s right panel, each vehicle can build an awareness zone
As demonstrated in field tests, the system could support Which- (AZ) using data received from everyone within its communica-
Road, Which-Lane, or even Where-in-Lane applications using tion range, and provide feedback to the driver. This concept is
the same basic architecture. central to V2V communication.
Why V2V? Location and situation awareness are key enablers V2V technologies eliminate two critical limitations of tra-
for almost all modern-day vehicle safety and convenience ditional sensor-based technologies. First, a single communica-
systems. These include GM OnStar, adaptive cruise-control tion sensor can communicate with enabled vehicles within its
28 GPS World | April 2009 www.gpsworld.com
Vehicle Integration | TRANSPORTATION
munication ranges of both V2V-equipped The across-distance plot (bottom) ity and positioning mode differences. In
vehicles (not to scale). shows the estimate available from the time between host and target system
System performance results from a full Which-Road and Which-Lane subsys- switch to RTK mode at intersection A
north-south loop are shown under two tems. Comparing with the reference, the (time 105 seconds), the Which-Road es-
scenarios: when both vehicles use the same Which-Lane subsystem output agrees with timate is completely driven by the absolute
onboard receiver type (type A), and when the reference solution at sub-meter level error difference of the vehicles. During
one uses type A and the other type B. at all times. However, the Which-Road these periods, errors experienced by one
Northbound Drive, Same Type. In the subsystem shows significant deviations vehicle are almost eliminated by intersec-
first test, both vehicles drove northbound at times, mostly driven by satellite visibil- tion RTK data (in RTK mode) while the
through the entire test area (FIGURE 7a).
The vehicles started in the same lane, with
the host following the target. The target
performed two lane-change maneuvers
during the run while the host remained in
the middle lane. During the lane change,
the target first changed to the left adjacent
lane, back to same lane formation with the
host, then changed to the right adjacent
lane to perform a similar maneuver.
For simplicity, the performance analy-
sis used the distances between vehicles,
in particular the between-vehicle across-
distance, for ease of validation. During a
same-lane formation drive, the between-
vehicle across-distance should be close
to zero, while in adjacent-lane drives, it
should approximately equal a lane width.
FIGURE 7b shows the absolute positioning
accuracy observed in both vehicles (top
plot) and the between-vehicle across-dis-
tance (bottom plot) as estimated by the
Which-Road and Which-Lane subsystems
using receiver type A in both vehicles. Also
shown is the reference across-distance gen-
erated by post-processing dual-frequency
data from the type B receivers.
The absolute positioning accuracy plot
shows three time periods in which the ac-
curacy improves from around 5 meters
RMS to better than 1 meter RMS. These
are the durations when the vehicles were
within the coverage of a V2I broadcast,
thus were receiving RTK data from inter-
sections. At intersection A, the target ve-
hicle position solution switches to RTK a
few seconds ahead of the host and remains
in RTK mode for around 35 seconds until
the vehicles depart from the intersection
coverage area. The absolute accuracy plot
shows that infrastructural support can
furnish Which-Lane accuracy. Outside
V2I coverage, both vehicles show typical
standalone GPS accuracy.
www.gpsworld.com April 2009 | GPS World 33
TRANSPORTATION | Vehicle Integration
5
u-blox 5
Under 1 second Time To First Fix
50 channels, 1 million correlators
SuperSense® -160 dBm sensitivity
Under 50 mW power consumption
Galileo-ready
your position is our focus Mobile World Congress 2008, Feb. 11-14, Barcelona, Spain
CeBIT 2008, March 4-9, Hannover, Germany
CTIA 2008, April 1-3, Las Vegas, NV, USA
Worst-Case Scenario
SBAS
Interference
Some applications using data from fixed or low relative
dynamic WAAS or other SBAS receivers may encounter
a tracking error due to C/A code cross-correlation.
L. Lestarquit, J-L. Issler, O. Nouvel, M. Sihrener, G. Lamontagne, J.C. Guay, R. Landry, O. Julien, C. Macabiau, M. Nouvel-Malicorne
G
PS C/A codes present cross-correlation peaks that pler between the two interfering signals lower than the receiver
can potentially cause false acquisitions. They can also code loop bandwidth, typically lower than 1 Hz. The signed
create tracking errors and C/N0 degradation, not in amplitude of the interfering peak is a function of the relative
traditional GPS tracking situations, but for applications where phase, and if the relative phase moves too much during the code-
signals are received with a low Doppler difference. loop integration period, the interfering peak (and the tracking
Several applications present a low relative Doppler, thus caus- error) will be filtered out. We refer to this condition as a Doppler
ing more frequent Doppler collision and this type of interference. collision, or as having a quasi-stationary code, because from the
One occurs with the use of ranging signals from geostationary receiver point of view, the received codes will not move relative
satellites (for example, satellite-based augmentation systems, or to one another. This means that the cross-correlation peak of
SBAS) when two or more of these satellites are in line-of-sight of the interfering signal will not move relative to the main auto-
a fixed- or low-dynamic receiver such as a control station, refer- correlation peak of the desired signal, and the interference will
ence station for differential positioning, or any ground user with result in a lasting bias.
little or no motion. The tracking error increases strongly with For common GPS applications, a Doppler collision smaller
more geostationary satellites — and multiple SBAS have been than 1 Hz is quite a rare phenomenon. Doppler due to satellite
and continue to be developed with overlapping interoperative motion is high (from ⳮ4.5 kHz to 4.5 kHz for standard use),
regions. Indeed, it is possible that all the signals will interfere at giving a low probability of such a Doppler collision; even if it
the same time, or sequentially, one after another. The SBAS C/A happens, it will last a few seconds. For common GPS applications,
code interference could also be seen for mobile receivers such as C/A code interference rarely leads to a noticeable tracking error.
airplanes, moving along a zero-relative Doppler line between two However, with the development of GNSS systems using geo-
SBAS satellite longitudes. stationary satellites such as EGNOS, WAAS, GAGAN, MSAS,
Pseudolite use can also induce low dynamic signals, in applica- BEIDOU-1, or IRNSS, the dynamics are much lower, in the
tions involving a pseudolite and a receiver onboard a geostationary range of a few meters/second, corresponding to a few 10s of Hz
satellite, and pseudolites and fixed- or low-dynamic receivers. at L1 frequency. Doppler collisions will occur twice a day and last
The interference mechanism is physically the same as for several minutes. Thus, interference due to PN-code code cross-
multipath error. Actually, multipath can be viewed as C/A code correlation will occur, and can cause false acquisition, C/N0 deg-
self-interference: the cross-correlation peak of the reflected signal radation, and tracking errors. Our results using data collected by
contributes to the correlation function and causes an offset of the WAAS receivers in Canada show the impact of several factors,
zero-crossing of the discrimination function, known as the track- including the navigation message, on the tracking error.
ing error. In C/A code interference, the cross-correlation peak of The current EGNOS inhibits the GEO ranging function, and
the interfering code creates the same effect. Normalized cross- therefore pseudorange measurement errors made on the GEO
correlation peak values are either Ⳮ63/1023 or ⳮ65/1023. The signal have no impact today on EGNOS performance. Moreover,
error can reach 18 meters for a half-chip spacing correlator when in more than 10 years of operation, including the EGNOS test
two quasi-stationary C/A codes are received with the same power bed, no message loss due to the C/A code interference has been
level. If the cross-correlation function of desired and interfering observed or reported on EGNOS. Our work here intends to:
codes has a secondary peak within the receiver’s chip spacing of 䡲 precisely characterize these interference tracking errors,
the relative code offset, there is potential for a tracking error. 䡲 identify the factors driving these errors, and
A necessary condition for such interference is a relative Dop- 䡲 propose mitigation techniques to reduce them.
www.gpsworld.com April 2009 | GPS World 37
SYSTEM DESIGN & TEST | Signal Processing
MULTIPATH
Retard relatif: -0.5, Puissance relative: 0
Retard relatif: -0.5, Puissance relative: 5
40 Retard relatif: -0.5, Puissance relative: 10
Retard relatif: 0.5, Puissance relative: 0
Retard relatif: 0.5, Puissance relative: 5
30 Retard relatif: 0.5, Puissance relative: 10
C/A CODE CROSS CORRELATION INTERFERENCE
-2 Tracking-Error Theory
-4 In the presence of an interfering signal, the receiver delay lock
-6 loop not only tracks the desired signal delay, but also the delay of
-8 the sum of the desired signal (index k) and all interfering signals
-10
-1.5 -1 -0.5 0 0.5 1 1.5
(index i ⫽ 1 to N).
Retard relatif (m) The received signal can thus be written as:
Å FIGURE 2 Theoretical tracking errors for pseudorange measurement
Analysis
The C/N0 measurements are clearly af-
fected by the cross-correlation interfer-
ence. Pairs of C/N0 degradation periods
are clearly visible during each Doppler
collision on June 19. These pairs are
explained by the slow shift of the cross-
correlation function. We measured deg-
radations of about 0.5 dB.
These pairs can be observed on the 3rd-
Å FIGURE 5 Highest WAAS message correlation zone
order derivative of the pseudorange. These
C/N0 degradations must be taken into ac-
count in the WAAS link budgets, and in
the computation of the L1 power increase
of SBAS modernization regarding L1. Fig-
ure 10 gives an example of such a 3rd-order
pseudorange derivative for PRN 135, for
the same period as the PRN 135 pseudor-
ange measurement in Figure 9.
Conclusions
This study set out to analyze in detail an
error affecting positioning performance
for users and applications using GEO
ranging measurements as offered today
by some SBAS. Such is the case for the
U.S. WAAS and the Japanese MSAS.
It is not the case for the European
Å FIGURE 6 Observation site near Kamloops, British Columbia, Canada
EGNOS, as EGNOS users are not
hour trial recording period, and a 48-hour correlation function. authorized to use these GEO pseudo-
continuous recording period. The results in FIGURE 9 are related to range measurements in the system’s cur-
FIGURE 6 shows the location, near Wal- the cross-correlation error residuals. PRN rent version.
hachin, 45 kilometers from Kamloops. 135 is received with 2.5 dB more power The positioning error mechanism ana-
FIGURE 7 shows a very long Doppler than PRN 138. Errors seem to comply lyzed here, due to interferences between
collision, lasting 1 hour, 45 minutes. The with theory. SBAS codes during Doppler collisions,
results in FIGURE 8 are related to the cross- The results in FIGURE 10 are related to is similar to the case of multipath errors.
Å FIGURE 7 SBAS Doppler evolution Å FIGURE 8 Relative Doppler between PRN 135 and 138
40 GPS World | April 2009 www.gpsworld.com
Signal Processing | SYSTEM DESIGN & TEST
Å FIGURE 9 PRN 135 and 138 cross-correlation function Å FIGURE 10 PRN 135 and 138 relative delay
were not observed at the end of our mea-
surement campaign, because one of the
two observed WAAS satellite made an
east-west maneuver during that period,
this having displaced the maximum in-
terference zone.
If n is the number of SBAS satellites in
a given region of the world, the number
of maximal interference zones covering
the region is n!/[2!⫻(n-2)!]. Four SBAS
over a region will produce six maximal in-
Å FIGURE 11 PRN 135 error w/o smoothing Å FIGURE 12 PRN 138 error w/o smoothing terference zones. Five SBAS satellites will
The most affected applications are those curate ranging. At a minimum, it would produce 10 maximal interference zones in
of fixed SBAS users, or ones provided with be necessary in case of GEO ranging to the region, and so on. These zones are not
low dynamics relative to the Doppler. detect the errors, and either to eliminate stable since the geostationary SBAS space-
Results show a possibility of mitigating all measurements made during Doppler craft maneuver sporadically.
the interference by applying a correction collision from the navigation solution, or The SBAS C/N0 degradation at L1 due
algorithm using the method suggested to adequately bound the associated errors to quasi-stationary C/A codes is due to a
by CNES: for instance, in all the user in the integrity computation. variation of the C component, rather than
WAAS receivers supposed to perform ac- The worst cases of Doppler collision a variation of the N0 component. We rec-
𰀲𰁁𰁋𰁏𰁎𰀀𰀍𰀀𰁌𰁅𰁁𰁄𰁉𰁎𰁇𰀀𰁔𰁈𰁅𰀀𰁗𰁁𰁙𰀀𰁗𰁉𰁔𰁈𰀀𰁃𰁒𰁙𰁓𰁔𰁁𰁌𰀀𰀀
𰀆𰀀𰀴𰀣𰀸𰀯𰀀𰁔𰁅𰁃𰁈𰁎𰁏𰁌𰁏𰁇𰁙𰀀𰁆𰁏𰁒𰀀𰁏𰁖𰁅𰁒𰀀𰀒𰀐𰀀𰁙𰁅𰁁𰁒𰁓𰀀
𰀯𰁕𰁔𰁓𰁔𰁁𰁎𰁄𰁉𰁎𰁇𰀀𰁐𰁅𰁒𰁆𰁏𰁒𰁍𰁁𰁎𰁃𰁅𰀀𰁉𰁎𰀀𰀧𰀰𰀳𰀌𰀀𰁃𰁏𰁍𰁍𰁕𰁎𰁉𰁃𰁁𰁔𰁉𰁏𰁎𰁓𰀌𰀀𰀀
𰁁𰁕𰁔𰁏𰁍𰁏𰁔𰁉𰁖𰁅𰀌𰀀𰁁𰁅𰁒𰁏𰁓𰁐𰁁𰁃𰁅𰀀𰁁𰁎𰁄𰀀𰁍𰁉𰁌𰁉𰁔𰁁𰁒𰁙𰀀𰁁𰁐𰁐𰁌𰁉𰁃𰁁𰁔𰁉𰁏𰁎𰁓𰀀𰁁𰁒𰁏𰁕𰁎𰁄𰀀𰁔𰁈𰁅𰀀𰁇𰁌𰁏𰁂𰁅𰀎
𰀳𰁕𰁐𰁅𰁒𰁉𰁏𰁒𰀀𰁐𰁈𰁁𰁓𰁅𰀀𰁎𰁏𰁉𰁓𰁅𰀀𰁐𰁅𰁒𰁆𰁏𰁒𰁍𰁁𰁎𰁃𰁅
𰀦𰁒𰁅𰁑𰁕𰁅𰁎𰁃𰁙𰀀𰁇𰁒𰁁𰁄𰁉𰁅𰁎𰁔𰀀𰀜𰀐𰀎𰀐𰀕𰁐𰁐𰁍𰀏𰂪𰀣
𰀋𰀏𰀍𰀀𰀐𰀎𰀕𰁐𰁐𰁍𰀀𰁆𰁒𰁅𰁑𰁕𰁅𰁎𰁃𰁙𰀀𰁓𰁔𰁁𰁂𰁉𰁌𰁉𰁔𰁙𰀀𰁏𰁖𰁅𰁒𰀀𰁗𰁉𰁄𰁅𰀀𰁔𰁅𰁍𰁐𰁅𰁒𰁁𰁔𰁕𰁒𰁅𰀀𰁒𰁁𰁎𰁇𰁅𰁓
𰁏𰁒𰁆𰁎𰀃𰁒𰁑𰀃𰁚𰁌𰁗𰁋𰀃𰀵𰁄𰁎𰁒𰁑
𰁉𰁌𰁑𰁇𰀃𰁗𰁋𰁈𰀃𰁅𰁈𰁖𰁗𰀃𰁓𰁕𰁒𰁇𰁘𰁆𰁗𰀃𰁗𰁒𰀃𰁖𰁘𰁌𰁗𰀃𰁜𰁒𰁘𰁕𰀃𰁑𰁈𰁈𰁇𰁖𰀃𰁄𰁗𰀝𰀃𰁚𰁚𰁚𰀑𰁕 𰁄 𰁎 𰁒𰁑𰀑𰁆𰁒 𰁐
Å FIGURE 13 Nav message cross-correlation, PRN 135 and 138 Å FIGURE 14 Nav message cross-correlation, PRN 135 and 138
Manufacturers
Tests used a NovAtel DL-4/OEM4 receiver (www.novatel.com ).
Å FIGURE 15 C/N0 of WAAS PRN 135 and 138
COME TO
ANNAPOLIS
GNSS Education
Annapolis, Maryland
June 15-19, 2009
GPS Operation for Engineers
and Technical Professionals
Course 356 (5 Days)
NavtechGPS
www.NavtechGPS.com
800.628.0885 or 703.256.8900
courses@navtechgps.com
42 GPS World | April 2009 www.gpsworld.com
Algorithms & Methods | INNOVATION
T
he main goal of this article is to the technique, along with its technical
describe the current performance limitations. We begin with a review of the
of what has become known as the current state of PPP, covering performance
precise point positioning (PPP) technique, and usage. We then discuss current
and to discuss the future potential of technical limitations of the approach,
1.0
0:00
0:15
0:30
0:45
1:00
1:15
1:30
1:45
2:00
2:15
2:30
2:45
3:00
infrastructure challenges, including the availability of precise
Time from session start (hours:minutes)
satellite orbits and clock offsets, precise orbit and clock prediction,
▲ FIGURE 1 St. John’s (STJO) PPP solution convergence on seven
real-time dissemination of predicted orbits and clocks, and sequential days.
reference frame realizations. Given the upcoming great changes
due to GPS modernization and the development of other global Performance Specifications
navigation satellite systems (GNSS), we would be remiss not to The standard metrics used here to describe the performance of
speculate on the potential significant positive impacts of these conventional PPP services are: accuracy, precision, convergence
added signals on future PPP performance. Finally, we end with a period, availability, and integrity. With PPP solutions showing
rather provocative discussion of the potential of PPP to perform very little in the way of biases — typically a few centimeters at
in a similar manner as the RTK technique. the most — there is very little difference between the accuracy
and precision metrics. In terms of north, east, and up component
Current Status accuracies at the 1-sigma level, PPP is able to provide few-
This section is designed to summarize current PPP performance centimeter-level results in static mode and decimeter-level results
using a number of metrics, and to set the technique’s impact in kinematic mode; both could be achieved in either post-mission
within the context of the wider field of positioning and naviga- analyses or in real time.
tion. What we mean by PPP is the state-space solution to the pro- The convergence period, namely the length of time required
cessing of pseudorange and carrier-phase measurements from a from a cold start to a decimeter-level position solution, is typi-
single GNSS receiver, utilizing satellite constellation precise orbits cally about 30 minutes under normal conditions and will be sig-
and clock offsets determined by separate means. Typically, a dual- nificantly longer before the position solution can converge to the
frequency GNSS receiver is used with dual-frequency code and few-centimeter level, if at all. This period is determined by the
phase measurements linearly combined to remove the first-order measurement strength of the observables for a GPS-only solution,
effect of ionospheric refraction. The real-valued carrier-phase am- the geometry of the problem, and the redundancy available for
biguity terms are estimated from the measurement model. The the estimation problem. Initial solutions rely almost exclusively
tropospheric refraction is also estimated, along with the receiver on noisy pseudorange measurements, the uncertainties of which
position and ambiguity parameters from the measurements. PPP are magnified via the ionosphere-free linear combination. The
using a single-frequency GNSS receiver has also been investigated availability of solutions is usually high, given that application
with great promise for certain applications. However, we will areas for this technique are open sky, continuously unobstructed
not discuss these further in this article; see Further Reading for environments. Otherwise PPP would typically not be used (at
publications reporting developments in single-frequency PPP its current level of development). Finally, the availability of in-
and other advances in the technique. To achieve the best position tegrity measures for the PPP solution is considered. Aside from
accuracy possible from PPP, effects such as carrier-phase wind-up, filter covariance estimates, quantitative quality measures of the
transmitter-antenna phase offset, solid Earth tides, and the con- obtained results are limited. For example, knowledge of biases
tribution of ocean tide loading must be corrected using models. in corrections such as precise satellite orbit and clock products,
Residual terms such as receiver noise and multipath are generally the potential for biases in estimated coordinates, and measure-
ignored or minimally handled using stochastic procedures. ment outliers are typically not considered (that is, not rigorously
A unique aspect of PPP is that it is an area of research being specified and accounted for) in solutions.
actively pursued by academia, government, and industry, in con-
cert and individually. As is typical, early development occurred in Use and Applications
research settings for scientific goals. Governments, as service pro- PPP can be used for the processing of static and kinematic data,
viders, have in some cases engaged in providing PPP services to the both in real time, if the dissemination mechanisms are in place to
public, given the socioeconomic benefits. Industry has embraced construct, transmit, receive and process precise satellite orbit and
and advanced the technology to better serve its clients. The results clock products, and in post-processing mode. The caveat for all
are: 1) rapid development and use of PPP in a variety of applica- such usage is that there needs to be uninterrupted GNSS signal
tion areas, and 2) significant overlap between the three sectors in availability, as loss of tracking lock on a minimum number of sat-
terms of research and development, and service models. The latter ellites requires processing filter re-initialization, resulting in tens
point will be discussed further in the infrastructure section. of minutes of greater than decimeter-resolution positioning, until
44 GPS World | April 2009 www.gpsworld.com
Algorithms & Methods | INNOVATION
1.0
phone 630.372.6800
toll-free 800.323.9122
website antenna.pctel.com
email sales.antenna@pctel.com
Toll Free Phone: 866-289-4777 • www.gpssource.com
Latitude
Latitude Longitude
Longitude Height
Height
3.0
2.4
ambiguity parameters are integers, which can render the position
1.8
Ge
Gett info rme d.
informed.
Network
Ne with industry
two rk with indus try leaders.
le ade rs .
Learn
Le what yo
arn what your
ur competition
c o mpe titio n isis doing.
do ing .
Attend
Atte the indus
nd the industry’s most
try’s mo s t respected
re s pe c te d conference.
c o nfe re nc e .
Details @ www.gps-wireless.com
www.gpsworld.com April 2009 | GPS World 47
INNOVATION | Algorithms & Methods
DRAPER LABORATORY
𰀦𰆌𰅽𰅵𰀃𰄐𰅽𰅶𰄐𰄞𰆉𰆚𰀃𰆚𰅽𰀃𰆉𰆌𰅽𰆚𰅽𰆚𰇇𰆉𰄞𰍕𰀃𰇁𰄞𰀃𰄞𰅶𰅐𰅝𰅶𰄞𰄞𰆌𰀃𰆐𰅽𰅯𰆵𰆟𰅽𰅶𰆐𰀃𰆚𰅽𰀃𰆉𰆌𰅽𰄏𰅯𰄞𰅵𰆐𰀃𰅝𰅶𰀃𰁞𰆉𰄂𰄐𰄞𰍕𰀃
𰀘𰄞𰄨𰄞𰅶𰆐𰄞𰍕𰀃𰀯𰅶𰆚𰄞𰅯𰅯𰅝𰅐𰄞𰅶𰄐𰄞𰍕𰀃𰀧𰄞𰅽𰆐𰆉𰄂𰆟𰄂𰅯𰍕𰀃𰀑𰅝𰅽𰅵𰄞𰄚𰅝𰄐𰄂𰅯𰍕𰀃𰄂𰅶𰄚𰀃𰀜𰅶𰄞𰆌𰅐𰇇𰀃𰁞𰇇𰆐𰆚𰄞𰅵𰆐𰍘
𰀒𰅚𰄂𰅯𰅯𰄞𰅶𰅐𰅝𰅶𰅐𰀃𰅽𰆉𰆉𰅽𰆌𰆚𰆵𰅶𰅝𰆟𰄞𰆐𰀃𰄨𰅽𰆌𰀃𰆋𰆵𰄂𰅯𰅝𰄮𰄞𰄚𰀃𰄐𰄂𰅶𰄚𰅝𰄚𰄂𰆚𰄞𰆐𰍗
𰀃 𰍻𰀃𰀧𰁗𰁞𰀃𰄂𰅶𰄚𰀃𰁅𰄂𰇀𰅝𰅐𰄂𰆟𰅽𰅶𰀃𰁤𰄂𰆐𰅬𰀃𰀾𰄞𰄂𰄚𰄞𰆌𰆐
𰀃 𰍻𰀃𰀧𰁗𰁞𰀃𰄂𰅶𰄚𰀃𰁅𰄂𰇀𰅝𰅐𰄂𰆟𰅽𰅶𰀃𰀜𰅶𰅐𰅝𰅶𰄞𰄞𰆌𰆐
𰀃 𰍻𰀃𰁚𰀦𰍬𰀒𰅽m𰅵𰆵𰅶𰅝𰄐𰄂𰆟𰅽𰅶𰆐𰀃𰀘𰄞𰆐𰅝𰅐𰅶𰄞𰆌𰆐𰀃𰄂𰅶𰄚𰀃𰀾𰄞𰄂𰄚𰀃𰀜𰅶𰅐𰅝𰅶𰄞𰄞𰆌𰆐
𰁙𰆵𰄂𰅯𰅝𰄮𰄞𰄚𰀃𰄂𰆉𰆉𰅯𰅝𰄐𰄂𰅶𰆚𰆐𰀃𰅵𰆵𰆐𰆚𰀃𰄏𰄞𰀃𰁨𰍘𰁞𰍘𰀃𰀒𰅝𰆟𰇌𰄞𰅶𰆐𰀃𰅽𰆌𰀃𰆉𰄞𰆌𰅵𰄂𰅶𰄞𰅶𰆚𰀃𰆌𰄞𰆐𰅝𰄚𰄞𰅶𰆚𰆐𰀃𰇁𰅝𰆚𰅚𰀃𰄂𰄏𰅝𰅯𰅝𰆚𰇇𰀃𰆚𰅽𰀃
𰆐𰄞𰄐𰆵𰆌𰄞𰀃𰀘𰅽𰀘𰀃𰄐𰅯𰄞𰄂𰆌𰄂𰅶𰄐𰄞𰍘𰀃𰀘𰆌𰄂𰆉𰄞𰆌𰀃𰅽𰄫𰄞𰆌𰆐𰀃𰄂𰅶𰀃𰄞𰇆𰄐𰄞𰅯𰅯𰄞𰅶𰆚𰀃𰄐𰅽𰅵𰆉𰄞𰅶𰆐𰄂𰆟𰅽𰅶𰀃𰄂𰅶𰄚𰀃𰄏𰄞𰅶𰄞𰄮𰆚𰆐𰀃𰆉𰄂𰄐𰅬𰄂𰅐𰄞𰍘
𰀄𰆉𰆉𰅯𰇇𰀃𰅽𰅶𰅯𰅝𰅶𰄞𰀃𰄂𰆚𰍗𰀃𰇁𰇁𰇁𰍘𰄚𰆌𰄂𰆉𰄞𰆌𰍘𰄐𰅽𰅵𰀃
𰀜𰀜𰍬𰀄𰀄𰀃𰄞𰅵𰆉𰅯𰅽𰇇𰄞𰆌 #*
Engineering Innovations
example, we demonstrated that combined 1. Can PPP algorithms, data and op-
GPS/GLONASS PPP greatly improves eration be improved to the point where
positioning accuracy and reduces the technique obtains the same level of
Compact antennas for handheld &
convergence time (see FIGURES 5 and 6). performance as RTK? And, if so, what Mobile GPS/SatCom
These improvements are dependent on specific improvements are required for
Quadrifilar
the enhanced level of satellite availability PPP to perform like RTK? Helical Antennas
and geometry for position determination. This result can be seen as a potential *10mm x 20mm
*Gain -1.3 dBic
Issues such as interoperability and final objective of PPP algorithm research *0.2 dB Axial Ratio
compatibility, however, must be addressed — and a recasting of the overall question. Microstrip
for successful integration of data from All of the measurement strength of the Patch Antennas
*Available in 5 sizes
hybrid navigation systems. undifferenced PPP observables is used *Gain 6.3 dBicv
to determine: 1) few-centimeter-level Now hiring RF & antenna engineers. BSEE
and US citizen prefered. Send resumes to
PPP versus RTK position estimates, 2) with a few seconds jobs@maxtena.com
Given the previous discussions about worth of measurements, 3) without the
1-877-MAXTENA
the capabilities and potential of PPP, an need for a reference station. The utility 1715 Pratt Drive, Suite 2800
Blacksburg, VA 24060, USA
obvious topic to conclude this article is a of such a solution would be significant www.MAXTENA.com
comparison of PPP and the industry-stan- (see below).
dard RTK technique, and to answer the The third characteristic (of removing
question: Can PPP ever replace RTK? To the need for reference stations) is the one research results related to ambiguity
address this overall issue, we pose a series that would make PPP so appealing as resolution of undifferenced observables.
of logically ordered questions. Some que- compared to RTK. The first characteristic The potential exists to isolate and
ries may be answered quite readily, while (of attaining centimeter-level positioning estimate initial fractional phase biases in
others can only be partially addressed. accuracy) may be possible, given recent order to isolate true integer ambiguities,
www.gpsworld.com April 2009 | GPS World 49
INNOVATION | Algorithms & Methods
3
Position error (centimeters)
2 East
North
1 Up
RTK, or Differential GPS?
0
We can speculate that PPP would f irst be used as a
-1
complementary solution to RTK in positioning and navigation
-2
-3
work. But as the approach gained acceptance, it could replace
6:40 7:10 7:40 RTK. The only caveat here is that a significant level of integrity
Time (hours:minutes)
▲ FIGURE 5 Simulated GPS-only PPP solution convergence period would have to be guaranteed along with accurate PPP solutions
and accuracy performance to gain industrial acceptance.
3
5. How would science, industry, and society be affected?
Position error (centimeters)