WILLIAM GIBBS

At least one possible future for the delivery of high-accuracy services from global navigation satellite systems is becoming clearer as time passes. A large number of networks will be available. They will be operated by private enterprise.

Surveyors and others in need of centimetre-level positioning have much to look forward to, although it will cost them.

Technically, it’s nothing new. It is now axiomatic that high-accuracy services over a wide range of locations can only be provided via networks of continuously operating reference stations. These CORS networks take differential corrections from a number of reference stations and use them to create models of the ionosphere, from which they can derive the position error at any point within the network.

In general, it seems the results are much better than is possible from a single differential base station. That is certainly true if the mobile receiver moves any distance during the survey.

To date, CORS services have been provided on a more or less experimental basis by state land authorities in NSW (CORSNet), Victoria (Vicmap Position), WA and Queensland (SunPos). Generally, these networks are centred on the state capital.

In the last year or so, attempts have been made in Victoria and NSW to extend the networks out to the state borders.

Other smaller networks are also evolving, such as the four-station Darwin network, two stations in Alice Springs, and two state-funded stations in Tasmania.

The frontrunner seems to be the Department of Sustainability and Environment, which has been funded to extend Vicmap Position all the way to South Australia. One Vicmap Position station has been established in Adelaide, giving continuous coverage across the Barossa.

Over the top of this is AuScope, a project to fund a national CORS network. It is run by Geoscience Australia. It would use its own stations, plus the state networks, to provide this service nationally to the science community.

However, it has to be said that, despite the investment they are making, government authorities are rather uncomfortable with the idea of running these networks as production systems for work-a-day farmers, surveyors or civil engineers.

So private enterprise has stepped into the breach. GPSNet Perth, established in 2006, was the first privately owned CORS network in Australia and now comprises 12 stations covering the city of Perth and its environs.

However, the model here is one of straightforward competition between GPSNet Perth and Landgate. Landgate has a network of stations which it has built and maintained for its own purposes. GPSNet Perth has a different network with different stations, covering part of the same area, which it runs for profit.

The problem with this as a national model is that it duplicates expensive hardware unnecessarily.

Two new companies have been formed recently that point to a different and more rational way of organising the CORS industry in the future.

SmartNet Aus, the Leica subsidiary that is building a network of reference stations across Australia, is making strides in the construction of the network, although it may well take five years to create national coverage.

Leica has already introduced the concept across Europe and in a number of other countries.

So far base stations have been established in South Australia and on the Gold Coast. Bob Morton, who is heading the development, says discussions are still underway with three of the state-owned systems, SunPos, Vicmap Position and CORSnet. Discussions about the federal AuScope network are also in progress with Geoscience Australia.

The other initiative is a private company called Checkpoint, which has been formed in Adelaide to launch a network of reference stations. It plans to sell subscriptions to the service to farmers, viticulturists and miners, as well as survey companies.

The first stage of the network was launched at the SSC conference in October. It consists of stations at Two Wells, Williamstown, Osborne, Hillcrest, Mount Barker and McLaren Vale. This is an area that brackets Adelaide, the surrounding hills and the Barossa Valley.

The second stage will consist of a further 14 stations surrounding the Spencer Gulf. The chief executive of the organisation, Desmond Elliott, says this is likely to be in place by July 2010.

Stage 3 will extend the network out to the Victorian border and to the west. It will also create links with Vicmap Position.

Both companies have a very similar business model, although they are targeting different types of market and have very different backgrounds.

At Smartnet Aus, Morton says users will buy a subscription to the network. The network operators then buy a nonexclusive right to data supply from individual reference stations, which will then be amalgamated into a network model and communicated back to users.

At Checkpoint, Elliott says his strategy is to use existing base stations created by farmers or miners. Signals from these stations will be included in the Checkpoint solution. Operators will receive payment for including their signal.

Checkpoint will be based on Novatel equipment, although it will be agnostic with respect to the brands used by users or base station proprietors. In a similar fashion, Smartnet will be a Leica shop, but it too will be constructed so it can give and receive data from any brand of equipment.

The first stations in the Checkpoint network have used Scorpion reference stations developed by GPSat Systems Australia in Melbourne.

Scorpion uses a Novatel choke ring antenna to retrieve signals from the Navstars. It broadcasts an NTRIPformat signal to the internet, from where it can be connected into the mobile phone network or into UHF radios to reach end users.

Checkpoint will built Scorpions into 2 metre cube containers for installation in the outback. They will use batteries charged by solar panels.

Checkpoint will built Scorpions into 2 metre cube containers for installation in the outback. They will use batteries charged by solar panels.

This is interesting because it implies that future networks could be rolled out very quickly, simply on the basis of commercial transactions. Future network operators might not own any stations at all.

It also implies that operators of base stations could make a nice business out of selling signals from their base station to as many network operators as possible.

If one takes the claims of the operators at face value, the networks will be fully interoperable. The brand of reference station or rover will be irrelevant. The major differentiators between networks will presumably be in the location and quality of the reference stations, and in the quality of the network solutions.

Location will matter because the length of baselines matters. However, given the advances in technology that seem possible in the next decade, it may not matter a lot. Researchers are now looking at doubling network baselines while holding accuracy constant. Baselines of several hundred kilometres may well be feasible.

The accuracy of the position of the reference station may also be a factor. In 2008 Chris Rizos from the University of NSW defined the quality of CORS networks in terms of the quality of their underlying base stations. Tier 1 includes the IGS stations (designed for scientific and geodetic purposes); Tier 2 the primary national geodetic network (designed for datum maintenance); and Tier 3 state-managed and/or private networks used for a range of applications.

The quality of the network solution will also make a difference. Even if the networks use the same base stations, each may give different results to a mobile in the same position.

It could well be that, in the end, you pay your money and you take your choice.
 
William Gibbs is a Sydney-based science writer.

Issue 44; December – January 2009