INTRODUCTION |
A variety of software packages are available for GPS phase data processing.
These packages have either been developed by universities
and government departments (for research uses, for internal operational
purposes, or for very high precision scientific applications), or by the
receiver manufacturers. Although the detailed
architecture of GPS software varies considerably from one package to another,
there are a number of functions that most comprehensive GPS surveying software
packages must carry out.
The major components of any package include software to aid pre-survey planning, decision making and reconnaissance (section 5.2 and section 5.3); to support field observations (including, increasingly, the high productivity survey methods such as "stop & go", "kinematic" and other such techniques, section 5.5), data pre-processing and checking (section 5.4 and section 7.3); single-baseline data processing (section 9.2); network adjustments and quality control (section 9.3 and section 10.4), and the transformation of the results into an already established geodetic reference system (section 11.1 and section 11.2).
Such a package does not exist! GPS manufacturers are continuously
refining the software that they offer with their products. There are, however,
a number of reasons why software refinement cannot be carried out to the
degree that is perhaps possible, and one of the most important has to do
with the dynamic nature of the GPS technology itself. Valuable development
resources are generally directed to the creation of new, or significantly
revised software, rather than enhancing existing software. Nevertheless,
it is worth considering what makes a "good" software system so
that purchasers of GPS technology can carefully evaluate the software component
of the total GPS surveying "package", as well as the hardware
component.
Various criteria can be used to evaluate software. The following is a list of some that may be appropriate.
Primary Criteria:
Secondary Criteria:
Accuracy and reliability
are dependent not only on the data processing algorithms that
are used, but also on the planning and field procedures, as well as the
hardware. Hence, if the field procedures result in the collection of poor
quality data, or an inadequate quantity of data, then no matter how sophisticated
the software, the accuracy and reliability of the baseline results will
suffer. However, accuracy (and to a lesser extent reliability) do directly
influence the selection of the mathematical models
and the processing strategies to be employed
for any given survey application.
Although efficient and fast processing is desirable, this will obviously be dependent upon the computer implementation, the algorithms used, the amount of data to be processed, and the accuracy required. High accuracy applications generally requires the flexibility to intervene at various stages to control the processing. For scientific or geodetic applications, commercial software is far from adequate. "Part per million" accuracy (standard surveying) applications place a higher priority on automation or ease-of-use, with minimal operator intervention.
Nowadays, an important marketing factor for a GPS "package" is its ability to support high productivity survey techniques (section 5.5.1). The field procedures are very different to those of the conventional static GPS surveying method. However, it is the innovative data processing algorithms that make possible the generation of high accuracy results in a fraction of the time required by the conventional GPS baseline surveying technique.
Efficient data flow implies a natural flow of data from raw to finished form in some logical and efficient sequence of operations, and consequently has a direct impact on the software structure. Once a software architecture has been specified, it then becomes possible to address the related secondary criteria of automation, data storage, and computer hardware requirements, etc. (The integration of a Data Management System is central to ensuring this desired outcome -- it is in this area that many of the early commercial software packages were weakest.)
Many of the above secondary criteria will be determined to a large part by the observation models and processing methodologies used, and hence will have different optimal realisations depending upon the application. It is also important that the software structure allow for flexibility in processing options, and be adaptable to changes as new models and methodologies are developed. This is particularly the case for real-time positioning implementations, or where centimetre-level positioning is required for mission critical applications such as machine guidance and control. Finally, the software should be portable to allow easy implementation in different computing environments. This may be difficult to achieve if the user interface is device dependent, and requires significant operator input. Hence the software structure, the operating system requirements, and DMS structure selected will require a similar trade-off having to be made between efficiency, speed, and automation on the one hand, and software portability and maintainability on the other.
For the following discussions, GPS data processing can be considered as consisting of essentially four distinct tasks:
Back To Chapter 4 Contents
/ Next Topic / Previous Topic
© Chris Rizos, SNAP-UNSW, 1999
