This project had as its aim to develop and test various
GPS positioning techniques which take advantage of a network of continuously
operating GPS base stations, linked by dedicated telephone lines to
a central processing facility. A number of graduate students and staffers
from SNAP and the Surveying and Mapping Laboratory, School of Civil
& Environmental Engineering, Nanyang Technological University (NTU),
Singapore, were involved. This close collaboration led to the initiative
to jointly sponsor a "GPS/GIS Showcase" in Singapore. The
first was held in November 1999. The second in November 2000.
NTU academic staffers Goh Pong Chai and Tor Yam Khoon,
together with Chris Rizos and Shaowei Han, jointly submitted a National
Science & Technology Board (NSTB) grant application, "Development
of an integrated multiple base station infrastructure to support concurrent
high precision differential GPS positioning applications", seeking
funding to establish a GPS test laboratory in Singapore based on four
high precision permanent GPS reference stations linked by high speed
data lines to NTU. In September 1998 funding was approved by the NSTB.
The Australian Research
Council has supported the also project (1999-2001).
Shaowei Han was a visitor at NTU during July 1999 (his
report). Ken Wong made a similar visit in September 1999 (his
report). Chris Rizos spent 5 months of his sabbatical at NTU
from September 1999 (his report).
During the period February-April 2000, NTU researcher
Dr. Xiaoming Chen, worked with SNAP/UNSW researchers (primarily Shaowei
Han and Horng-Yue Chen)
to develop the basic processing 'engine' for the analysis of the reference
receiver network data. This 'engine' is software that lies at the heart
of network-based positioning, and must process data from multiple receivers
in real-time. A first demonstration of the capability was made in late
2000.
Database management and QC lies at the heart of such a
system. Preliminary work was carried
out in 1999 by Ken Wong to develop the necessary data handling procedures
to support both real-time applications, as well as web-based processing
and data archiving requirements.
The SIMRSN has operated for several years. In 2003 it
was decided to establish a 'mirror' site in Sydney. This real-time network
is known as SydNet.
Further algorithm development will be undertaken at UNSW from 2003 onwards.

Motivation: Use of multiple GPS reference
stations allows for "network-based" positioning solutions, rather than
the conventional single baseline (two receivers: one reference, the
other the user receiver) approach. Network-based GPS is of course the
mode used in ultra high accuracy GPS geodesy applications such as the
determination ground fault motion through the integrated processing
of data collected by a network of simultaneously operating GPS receivers.
The global network of GPS tracking stations comprising the "core" network
of the International GPS Service (IGS) is an example of such a multi-reference
receiver network. However, the data is not streamed to a computation
centre in real-time. Such a mode is also used in the pseudo-range-based
Wide Area Differential GPS (WADGPS) real-time services such as those
provided by organisations such as FUGRO (the "Omnistar" service). A
commercial version has been launched by Trimble, known as the 'Virtual
Reference System' (VRS), followed by Leica's version several years later.
Research Challenge: The multiple GPS reference
station network such as the one established in Singapore is similar
in concept to the WADGPS services. In both cases there is a separation
between the reference station network infrastructure and the user(s).
The reference station network is linked by dedicated communications
links to a central facility, where all the reference receiver data are
processed on a continuous basis, and "network correction messages" are
generated. These correction messages (not unlike the RTCM messages in
the case of single reference station DGPS implementations) are immediately
transmitted to users via appropriate data channels. The crucial difference
is that the network-based GPS technique uses carrier phase observations,
and therefore is intended to address high accuracy surveying and navigation
applications.
Various combinations of static and kinematic positioning
are possible, in either real-time (if there is a data channel available
to the user) or post-processed mode. Furthermore, users need not own
and operate the reference station network infrastructure. The latter
characteristic is significant, as it permits the development of network
services to the user, to support high accuracy applications, without
the need for a user to invest in base station receivers.
Building Monitoring Application: An example
of how an 'off-the-shelf' system can be adapted to improve performance
by using the reference station network is building monitoring.
A joint NTU-UNSW project (commenced in early 2000) requires the continuous
monitoring of Singapore's tallest building, the Republic Plaza Building,
using dual-frequency RTK-capable GPS receivers. The resultant time series
of positioning solutions will be analysed to detect vibrations of the
building structure caused by wind effects. Filtering algorithms were
developed as part of Clement
Ogaja's PhD project. Download
paper here describing monitoring system and time series analysis
methodology.
L-Commerce Products & Services: The
Singapore facility will permit research and development to be carried
out at the infrastructure level (reference station data processing,
network correction message generation, etc.), the data coms link level
(via a variety of technologies including GSM, Digital Audio Broadcast,
and packet radio services), and at the user/application level. A paper
that discusses how the 'future of GPS surveying' could be influenced
by the provision of 'services' such as those to be encouraged that take
advantage of this network facility has been prepared (a
PDF file can be downloaded here). Opportunities also exist to
address so-called 'L-commerce' applications, loosely defined as be 'position-dependent
services' delivered via consumer devices such as 'smartphones' or PDAs.
The Singapore-based project therefore offers unique opportunities to
address a wide range of positioning needs, from 'niche' applications
such as surveying, to mass market 'location-based-services' or 'L-commerce'
products and services. This research will be continued on the SydNet
GPS network.
Some Relevant
Publications: