Journal of Global Positioning Systems (2004)
Vol. 3, No. 1-2: 296-301
Development of SydNET Permanent Real-time GPS Network
C. Rizos, T.S. Yan
School of Surveying & Spatial Information Systems, University of New South Wales, Sydney, NSW 2052, Australia
e-mail: thomas.yan@unsw.edu.au Tel: + 61 2 9385 4189; Fax: +61 2 9313 7493
D.A. Kinlyside
Land and Property Information, Department of Lands, Bathurst, NSW 2795, Australia
e-mail: doug.kinlyside@lands.nsw.gov.au Tel: +61 2 6332 8372; Fax: +61 2 6332 8479
Received: 15 Nov 2004 / Accepted: 3 Feb 2005
Abstract. Over the past few years, there has been
substantial growth in multiple-reference-station networks
used to overcome the limitations of standard real-time
kinematic (RTK) systems. SydNET is a project to
establish a permanent real-time GPS network in the
Sydney basin area providing Network-RTK support to
users in the area. SydNET is being developed by NSW
Department of Lands in partnership with the School of
Surveying & SIS at the University of New South Wales.
This paper presents recent developments of the SydNET
network. Preliminary test results will be presented which
will show the network’s performance, achievable
accuracy. It will outline the SydNET system, its
operation, current status and vision of future development
as a high precision positioning service infrastructure.
Key words: RTK, Network-RTK, GPS networks, CORS
infrastructure
1 Introduction
1.1 Background
The standard mode of precise differential GPS
positioning is for one reference receiver to be located at a
base station whose 3D coordinates are known in a
geocentric reference frame so that the second receiver’s
coordinates are determined relative to this reference
receiver. This is the principle underlying pseudorange-
based differential GPS (or DGPS for short) techniques.
To achieve high accuracy, carrier phase data must be
used but this comes at a cost of system complexity
because the measurements are ambiguous. Therefore
Ambiguity Resolution (AR) algorithms must be
incorporated as an integral part of the data processing
software. Such high accuracy techniques are the result of
progressive R&D innovations which have been
subsequently implemented by the GPS manufacturers in
their top-of-the-line ‘GPS Surveying’ products e.g., Rizos
(2002a).
Over the last decade several significant developments
have resulted in this high accuracy performance also
being available in real-time. That is, immediately
following the making of measurements, and after the data
from the reference receiver has been transmitted to the
field receiver for processing via some data
communication links (e.g., VHF or UHF radio, mobile
phone, FM radio sub-carrier or satellite link), accurate
positions are produced in the field. Real-time precise
positioning is then possible when the GPS receiver is in
motion. These systems are commonly referred to as real-
time kinematic or RTK systems and make feasible the use
of GPS for many time-critical applications such as
engineering surveying, GPS-guided
earthworks/excavations, machine control and other high
precision navigation applications e.g., Lachapelle et al.
(2002).
The limitation of single-base RTK is the distance
between reference receiver and the user receiver due to
distance-dependent biases occurring such as orbit error
and ionospheric and tropospheric signal refraction. This
has restricted the inter-receiver distance to 10km or less if
very rapid Ambiguity Resolution (AR) is desired (ie less
than a few seconds).
Wide Area Differential GPS (WADGPS) and the Wide
Area Augmentation System (WAAS) on the other hand,
Rizos, Kinlyside, Yan: Development of SydNET Permanent Real-time GPS Network 297
use a network of base stations separated by hundreds of
kilometres over a wide geographic area. The
measurement biases can be modelled and corrected at the
users’ receiver and therefore the positioning accuracy will
be almost independent of the inter-receiver distance.
However, these are predominately pseudorange-based
systems intended to deliver accuracies at the metre to
sub-metre level.
Continuously Operating Reference Stations (CORS) have
been deployed to support very high accuracy geodetic
applications since the 1980s (Evans et al., 2002).
Geodetic techniques are by their very nature ‘multi-
station’, taking advantage of the geometric strength,
reference datum stability (and redundancy) afforded by
network-based positioning. Such CORS networks have
been deployed globally, as well as in geodynamic ‘hot
spots’ like Japan and Southern California where there is
significant tectonic motion (Ibid, 2002).
In Europe, as in many other countries, countrywide
‘active control stations’ have been established, consisting
of CORS that collect data specifically for survey and
mapping applications. In Australia, the state-wide
Victorian CORS network, GPSNet serves the same
purpose. Until recently however, such CORS networks
have contributed to improving surveying productivity by
obviating the need for GPS surveyors to operate a
reference receiver. They have not been used in an optimal
manner to address the distance constraint of single-base
positioning (real-time or post-mission) in the same way
that the WADGPS/WAAS techniques have done so far
for pseudorange-based positioning.
1.2 Network-based positioning
Rapid static and kinematic GPS surveying techniques can
become more productive by taking advantage of CORS
networks in such network-based implementations as
Network-RTK or more generally Network-based
Positioning (Lachapelle et al., 2002).
CORS networks support multi-station (e.g. geodetic)
processing of data from reference receivers
simultaneously with the user receiver data. These days,
this is typically achieved by a web-based service such as
AUSPOS
(http://www.auslig.gov.au/geodesy/sgc/wwwgps/), which
requires the user to upload their data to the web-engine,
subsequently returning the results of the processing to the
user. AUSPOS is not a real-time service and currently
only supports static positioning for occupations of several
hours or more using dual-frequency user receivers. GPS
surveyors use these online processing services to
establish high-order geodetic control but the services are
obviously unsuitable for high productivity engineering-
type surveys. AUSPOS and similar services rely on data
collected and archived by the International GPS Service
(http://igscb.jpl.nasa.gov), hence inter-receiver distances
are many hundreds (even thousands) of kilometres. The
Australian Regional GPS Network
(http://www.auslig.gov.au/geodesy/argn) is an example of
a sparse network that augments the IGS global network
and provides data to web-engines such as AUSPOS.
To address applications other than geodesy/geodynamics,
many countries and states have established CORS
networks that collect data for users to subsequently
access and process themselves. This is an important
distinction; as such networks only provide ‘passive’
services such as data downloads of RINEX-formatted
measurement files. As with ‘geodetic’ CORS networks,
user only needs to operate one GPS receiver. However,
because the survey user must process data using software
typically provided by the GPS manufacturer, and rapid
GPS survey techniques are used (e.g. kinematic, rapid
static, ‘stop-and-go’, etc. – Rizos, 2002a), the distance
between the user receiver and the closest reference
receiver must be less than the maximum recommended
for GPS surveying applications. This is less than 10km
for very rapid AR and typically 20-30km for rapid static
techniques.
The Hong Kong GPS Network (Kwok, 2002) is an
example of a CORS network with a density of base
stations that a user is always within 10km of a reference
receiver (and usually two, to permit checking). On the
other hand, the state-wide GPSNet
(http://www.land.vic.gov.au/GPSnet/) established in
Victoria, is a typical example of a ‘passive’ CORS
network, with base station spacing of between 50 and
100km. In order to upgrade such a CORS network to real-
time operations would require the implementation of a
Network-RTK system if no user were to be
disadvantaged by being more than 10km from a base
station.
Several European countries have upgraded their CORS
networks to implement RTK. In some cases, such as in
Denmark where the density of base stations is high, of the
order of 10-20km station spacing, it is possible to use
standard single-base RTK techniques (Leica, 2003,
personal communication). In Germany, the Satellite
Positioning Network (SAPOS) (Elsner, 1996) has been
upgraded in recent years to offer a Network-RTK service
across all German states. This is a model that is likely to
be followed by other ‘passive’ CORS networks as they
upgrade to real-time operations.
1.3 Network-RTK concept
Network-RTK is the logical outcome of the continuous
search for a GPS positioning technique that challenges
the current constraints of single-base RTK, namely the
298 Journal of Global Positioning Systems
need to be within 10km of the base station if the highest
performance is to be achieved. It is a centimetre-
accuracy, real-time, carrier phase-based positioning
technique capable of operating over inter-receiver
distances up to many tens of kilometres with equivalent
performance to current single-base RTK systems. The
most crucial characteristic of contemporary RTK
techniques that must be preserved is very rapid time to
ambiguity resolution (AR), measured in seconds. Hence
the base stations must be deployed in a pattern dense
enough to model distance-dependent errors to such
accuracy that residual double-differenced carrier phase
observable errors can be ignored in the context of such
rapid AR (Rizos, 2002b).
Network-RTK requires a data processing ‘engine’ with
the capability to resolve the integer ambiguities between
the static reference receivers that make up the CORS
network. The ‘engine’ must be capable of handling
double-differenced data from receivers 50-100km apart,
operate in real-time, instantaneously for all satellites at
elevation cut-off angles down to a couple of degrees
(even with high noise data that is vulnerable to a higher
multi-path disturbance). The Network-RTK correction
messages can then be generated.
The benefits of the Network-RTK messages (as opposed
to standard RTK messages) are:
Elimination of orbit bias and ionosphere delay.
Reduction of troposphere delay, multi-path
disturbance and observation noise.
RTK can be extended to what might be
considered ‘medium-range’ baselines (up
100km).
Low-cost single-frequency receivers can be used
for RTK and rapid static positioning.
Very high accuracy applications using low-cost
GPS receivers (e.g. deformation monitoring,
geodetic control network, etc.) are possible.
Improve the accuracy, reliability, integrity,
productivity and capacity of GPS positioning.
In addition to the data processing engine, the Network-
RTK system needs to have a data management system
and a data communication system. It needs to manage
corrections generated in real-time, the raw measurement
data, multi-path template for each reference stations (for
multi-path mitigation), ultra-rapid IGS orbits, etc. There
are two aspects to the data communication system: (a)
between the master control station (MCS - where the data
processing engine and data archive are located) and the
various reference stations, and (b) communication
between the MCS and users. From the Network-RTK
implementation point of view, there are three possible
architectures (Rizos, 2002b): (1) generation of the Virtual
Reference Station and its corrections, (2) generation and
broadcast of an Area Correction Model, or (3) broadcast
the raw data from all of the reference stations. These are
briefly described below.
2.1.1 Virtual Reference Station (VRS)
At the MCS server, the VRS can be generated
and the RTCM 20/21 message created and
transmitted once the server knows the position
of the roving user. There is no further request
from the roving user if the rover supports RTCM
20/21 format, except that the user needs to send
their location to the server.
Two-way communication is required, with the
user informing the server where they are, and the
server continuously sending data to the user for
RTK applications.
There are some limitations on the number of
simultaneous users accessing the VRS service
due to server capacity.
This configuration has been used by
Trimble/Terrasat in their commercial product,
The Trimble Virtual Reference Station (Vollath
et al., 2000).
2.1.2 Area correction model broadcasting
At the MCS server, the corrections, e.g.
dispersive and non-dispersive atmospheric
correction terms or carrier phase measurement
residuals for each satellite at each reference
stations, will be generated using data from the
CORS network.
The corrections can be used to generate an
interpolation model or the VRS at the user end.
The correction generation algorithms can be
different.
One-way communication is sufficient and there
is no limit on the number of users. This requires
a new data format, and the volume of
transmitted data is more than in the case of a
single reference station.
This configuration has been proposed as a
Network-RTK RTCM format by Leica and
Geo++, and will be implemented in RTCM
version 3 (Han, 2003, personal communication).
Rizos, Kinlyside, Yan: Development of SydNET Permanent Real-time GPS Network 299
2.1.3 Raw data broadcasting
Broadcast raw measurements (CMR or RTCM
18/19 message format) from either the MCS
server or from the multiple reference stations
individually.
Generate the VRS, or corrections, at the user
site. The computation load is therefore shifted to
the user.
This requires a new data format.
One-way communication is sufficient.
There is no limit on the number of users.
A discussion of the pros and cons of each type of
implementation is beyond the scope of this paper. Tests
will need to be conducted to determine which of these is
best suited for the type of applications that will be
addressed by the network service.
1.4 The Singapore Integrated Multiple Reference
Station Network (SIMRSN)
Due to the complexity and cost (typically between $30-
$50,000 per station) involved in establishing CORS
networks, the data links and the data
processing/management servers at the control centre,
there have been comparatively few university-based
Network-RTK systems established to support research.
During the last few years, to the best of the authors’
knowledge, only the Singapore Integrated Multiple
Reference Station Network (SIMRSN) has been
operating both as a research facility and an operational
Network-RTK service for the benefit of GPS surveyors.
The SIMRSN is a joint research and development
initiative between the Surveying and Mapping
Laboratory, Nanyang Technological University (NTU),
Singapore; the School of Surveying & Spatial
Information Systems at the University of New South
Wales (UNSW); and the Singapore Land Authority
(currently the main user) (Chen et al., 2000).
The SIMRSN consists of five CORS, connected by
dedicated ISDN data lines to the control centre at NTU. It
is a high quality and multi-functional network designed to
serve the various needs of real-time precise positioning,
such as surveying, civil engineering, precise navigation,
road pricing, etc. (Chen et al., 2000). The SIMRSN also
services off-line users, who can access archived RINEX
data files via the Internet. The inter-station distances are
of the order of several tens of kilometres at most.
However, tests conducted in 2001 have shown that even a
network with such comparatively short baselines had
difficulty in modelling the disturbed ionosphere in
equatorial regions during the last solar maximum period
of the 11 year sunspot cycle (Hu et al., 2002). Unique
facilities such as SIMRSN can therefore act as a test bed
for network-based positioning techniques. The SIMRSN
model of a network that is both a research facility and an
operational network service for users is being adopted for
the SydNET network being established in the Sydney
metropolitan area.
2 SydNET CORS Network
2.1 Introduction
The SydNET real-time CORS network is being
established with network-based positioning capability
from the very start, including Network-RTK. The project
is funded by the NSW Department of Lands (Lands) as
an initiative of State Government infrastructure. Lands
has been active in using GPS for a variety of surveying
and mapping applications for over a decade (Kinlyside,
1999, 1995, 1993). The development of a Network-RTK
system for the state’s largest capital city is a natural and
logical extension of the organisation’s previous and
current involvement in GPS applications for surveying
and geodesy.
SydNET is also an important research facility for the
Cooperative Research Centre for Spatial Information
(http://www.spatialinfocrc.org/programs.html). It will be
available for testing various network-based positioning
techniques, both commercial products and those
developed by research organisations. Additionally,
SydNET can also be utilised for experiments on non-
positioning applications such as ‘GPS meteorology’.
2.2 Project Management and Structure
As with most other current projects within Lands,
SydNET Project is managed using the PRINCE2 project
management methodology. Major authorisations for the
project are made through a project board which currently
comprises two executives from Lands, one external
member from UNSW and one external member from the
Roads & Traffic Authority (RTA).
The School of Surveying & SIS at the University of New
South Wales is the main IP supplier and development
contractor. The first phase of SydNET is only servicing
the Sydney basin region - an area of approximately
100x100km - but it is planned for expansion over time to
cover other areas in NSW. In this initial phase, SydNET
is implemented using the SIMRSN Network-RTK
algorithms which support VRS-style Network-RTK and
provide an online service for RINEX data download.
300 Journal of Global Positioning Systems
2.3 Locations and Coverage
In 2003, Lands approached RailCorp with the idea of
utilising their fibre-optic network in order to provide the
communication links between receivers at the reference
stations and servers at the Network Control Centre
(NCC). RailCorp is a state-owned corporation which
provides passenger rail network throughout NSW.
RailCorp has fibre-optic network installed extensively
throughout their electrified railway network.
As many as eighteen sites on RailCorp’s network were
inspected for feasibility of installing a GPS reference
station with the intention of providing coverage over
Sydney basin area based on 15 to 20km spacing. Seven
sites are subsequently chosen because they provide the
best possible sky view, least radio interference, safety and
suitable infrastructure for installing a permanent GPS
antenna and receiver. These sites are located in
Chippendale, Villawood, Waterfall, Mulgrave,
Springwood, Cowan and Menangle. At the time of
writing, six stations have been installed at these locations.
Menangle is still to be constructed.
Figure 1 Location of SydNET sites and their coverage
2.4 Reference station hardware
Each reference station consists of a dual-frequency
geodetic grade GPS receiver with choke-rings antenna
and a device converting serial data streams into TCP/IP
packets over Ethernet. SydNET is using existing GPS
receivers belonging to Lands and UNSW, with the
intention of reducing the amount of initial capital
investment required. Currently, three Trimble 4000SSe
receivers from Lands are being deployed in SydNET
stations with a number of Leica System 500 receivers
from UNSW available for testing and ad-hoc stations.
Four new Ashtech µZ-CGRS receivers have also been
procured and deployed to augment the network. The
receivers will be upgraded as new GPS signals and the
Galileo GNSS signals become available to users.
2.5 SydNET communication link
Data from the reference stations is transmitted via
RailCorp’s network back to the NCC in TCP/IP packets
over Ethernet. There is a 10 Mbps physical link to
RailCorp’s network at each station. This is more than
sufficient for the current data stream which is in the range
of 20 to 30 kbps. Currently there is one main data stream
of raw GPS measurements at a rate of 1Hz. With
significant remaining bandwidth available, it is possible
to add more data streams or equipments for testing and
research purposes. In some of the stations, multi-ports
serial converters are being installed to enable different
streams of data for operational, research and signalling
purposes.
2.6. SydNET Network Control Centre
The Network Control Centre (NCC) for SydNET is
hosted by the Australian Centre for Advanced Computing
and Communications (ac3) located at the Australian
Technology Park, Redfern NSW. Ac3 provides a
professionally managed, premium facility that was
purpose built for high availability, is highly secure and
highly connected to the Internet backbones, including a
connection to RailCorp’s fibre-optic network.
Data streams from SydNET reference stations are
aggregated at ac3. A media converter was installed by
RailCorp in Lands equipments cabinet which connects
RailCorp’s fibre-optic network to Lands’ network.
Measurements data from all SydNET stations are
processed by SydNET servers to provide real-time
correction and also post-processing data.
Traditionally, correction data is transmitted to users via
radio on a specifically allocated frequency. This requires
specialised hardware both on the service provider’s side
and on the users’ side and also specific radio spectrum
license. A more advanced technique is to use GSM
mobile network but this still requires service provider to
invest in expensive modem banks and routers. A dial-in
system such as GSM has limited number of lines and
hence users at the same time. Another disadvantage is the
cost associated with using mobile network.
2.7 Real-time data distribution
Real-time correction data from SydNET is distributed via
the Internet. The choice is then given to users to arrange
access to the data on the Internet. A suggested method is
Rizos, Kinlyside, Yan: Development of SydNET Permanent Real-time GPS Network 301
by using General Packet Radio System (GPRS) service
which is available in most part of Sydney. With a suitable
GPRS-enabled device, users can access the data and
connect it to their GPS receiver. For testing and internal
use, Lands and UNSW use either a Compaq iPAQ PDA
with a GPRS jacket or an O2 XDA with built-in GPRS
phone. Both devices run Pocket PC operating system.
Client software has been developed to access SydNET
server and retrieve correction data. This data is then
transmitted on the device’s serial port which is connected
to the GPS receiver in the field.
Other alternatives to GPRS network are CDMA1x and
3G which are both available in Sydney as well. Similar to
GPRS, users can get either an add-on CDMA 1X card or
a PDA with built-in CDMA 1X capability such as the i-
mate PDA2K.
3 Conclusions
The proliferation of CORS networks at all scales, global,
national, state and local, will be a challenge to
organisations that seek the implementation of common
standards of service, and those that wish to see a seamless
network-based positioning capability across all networks.
The integration of networks with different operators, and
different functionalities, is an added challenge. In
Australia, there is the opportunity to address these
challenges at the national and state level through such
initiatives as the ‘network research’ to be undertaken by
the CRC-SI (http://www.spatialinfocrc.org).
The SydNET network may be considered a ‘second
generation’ CORS network, as it will be established with
network-based positioning capability from the very start.
The physical infrastructure, the communication links and
the database are all controlled by one agency, the NSW
Department of Lands. By providing such a framework,
new reference receivers will be able to be connected to
this ‘backbone’ on an ad hoc basis. The SydNET CORS
network is the first step in ultimately replacing NSW’s
primary geodetic network of trig stations with an
extensive network of ‘active control stations’.
References
Chen X.; Han S.; Rizos C.; Goh P.C. (2000): Improving real-
time positioning efficiency using the Singapore
Integrated Multiple Reference Station Network
(SIMRSN), 13th Int. Tech. Meeting of the Satellite Div. of
the U.S. Institute of Navigation, Salt Lake City, Utah, 19-
22 September, 9-18.
Elsner C. (1996): Real-time differential GPS: The German
approach, GIM Int. Journal for Geomatics, 10(5), 76-79.
Evans A.G.; Swift E.R.; Cunningham J.P.; Hill R.W.; Blewitt
G.; Yunck T.P.; Lichten SM, Hatch RR, Malys S, Bossler J
(2002): The Global Positioning System Geodesy Odyssey,
Navigation, 49(1), 7-34.
Hu G.; Khoo H.S.; Goh P.C.; Law C.L. (2002): Performance of
Singapore Integrated Multiple Reference Station
Network (SIMRSN) for RTK positioning, GPS Solutions,
6(1-2), 65-71.
Kwok S. (2002): The Hong Kong GPS network and reference
stations, Journal of Geopsatial Engineering, 2(2), 57-65.
Kinlyside D.; Jones G. (1999): DGPS Applications in the Land
Information Centre, Procs 4th Int. Symp. on Satellite
Navigation Technology & Applications, Brisbane, QLD,
20-23 July 1999.
Kinlyside D.; Bannister M.; (1995): Using OmniSTAR and
GPS in Outback NSW, Procs Satellite Navigation
Technology Conference, UNSW, Brisbane, QLD, 26-28
June 1995.
Kinlyside D.; Dickson G.; (1993): DGPS Activities of the Land
Information Centre, Procs Satellite Navigation
Technology Conference, UNSW, Kensington, NSW, 19-21
July 1993.
Lachapelle G.; Ryan S.; Rizos C. (2002): Servicing the GPS
user, Chapter 14 in Manual of Geospatial Science and
Technology, J. Bossler, J. Jenson, R. McMaster & C. Rizos
(eds.), Taylor & Francis Inc., 201-215.
Rizos C. (2002a): Making sense of the GPS techniques,
Chapter 11 in Manual of Geospatial Science and
Technology, J. Bossler, J. Jenson, R. McMaster & C. Rizos
(eds.), Taylor & Francis Inc., 146-161.
Rizos C. (2002b): Network RTK research and implementation:
A geodetic perspective, Journal of Global Positioning
Systems, 1(2), 144-150.
Vollath U.; Buecherl A.; Landau H.; Pagels C.; Wagner B.
(2000): Multi-base RTK positioning using Virtual
Reference Stations, 13th Tech. Meeting of the Satellite
Div. of the U.S. Institute of Navigation, Salt Lake City,
Utah, 19-22 September 2000, 123-131.