See current research theme 'Algorithms & Software'

Scenarios for Implementing Low-Cost GPS Surveying
& Stochastic Modelling Issues

Considerable R&D has been invested by instrument manufacturers to overcome those GPS technology problems which are considered to be unnecessarily restrictive for the surveying industry. Present state-of-the-art systems are: (a) able to be used in kinematic mode, (b) require comparatively short observation times, and (c) capable of real-time operation through the provision of a communication link between the two GPS receivers. However, a pair of such GPS receivers (and associated software), capable of operating in the real-time, kinematic mode typically over US$40K. At this price many surveyors are unable or unwilling to embrace the GPS technology for their day-to-day applications.

This project had as its aim to design, and implement, several surveying system configurations capable of delivering accuracies in the centimetre-decimetre range for significantly lower cost than the present systems. Former graduate student 'Chon' Satirapod made investigations in this regard for his PhD. One focus of research is the improvement of the stochastic and functional models for single-frequency baseline processing. This has been a recurring research theme of the SNAP group (see Theme 1).

There are several ways in which "low-cost" can be defined. For example, the use of a single user receiver, and taking advantage of a GPS reference receiver network service. This was discussed in a paper on the 'future of GPS surveying (download a PDF file here). Another possibility is to promote the use of inexpensive single-frequency GPS receivers in place of dual-frequency instruments. However, it must be demonstrated that the 'productivity' of such systems is not so low as to make them uneconomic. In this regard, current research is focussed on improving the stochastic model for single-frequency measurements, in the hope of improving the 'robustness' of ambiguity resolution for 'rapid static' modes of baseline determination. (Ambiguity resolution being the crucial operation, that is able to be much more reliably completed using dual-frequency data than single-frequency data, for very short observation spans.)

One scenario still being investigated is based on the use of low-cost, single-frequency GPS boardsets, used in conjunction with a communications link (e.g. mobile phone) to a web server connected to one or more GPS base stations. One example is to take advantage of multiple reference station networks (such as established in Singapore and in Sydney) to improve the performance of 'rapid static' techniques using single-frequency GPS receivers. Single-frequency RTK software has been developed (Theme 1), and is currently being refined on the SydNet GPS network.

Publications related to this topic: