By Craig Liddell

August 16, 2002

Frustrated by poor Internet access in regional areas, a NZ project is developing a wireless, last mile alternative.

The Waikato Wireless Project aims to develop a new platform that can be used to deploy future generations of high bandwidth (>10Mbps) wireless networks in rural and remote parts of New Zealand (NZ).

“It is widely recognised that telecommunications access in rural areas of New Zealand is poor,” says Dr Murray Pearson, senior lecturer, Department of Computer Science, University of Waikato, New Zealand. The project leader adds, “it’s increasingly becoming a barrier to businesses, including farmers, competing in the global economy.”

The WAND (Waikato Applied Network Dynamics) networking research group in the Department of Computer Science at the University of Waikato started the CRCnet (Connecting Remote Communities Network).

The group spoke to a number of rural community groups including schools who were frustrated by trying to get a decent Internet connection. “In many rural areas it is difficult to get any form of connection,” says Pearson. “The phone lines are low quality and things like poorly earthed electric fences can also cause real problems.”

The research, he says, “will explore the use of high performance point-to-point wireless protocols and the deployment of a redundant mesh structure. Today’s network management protocols are not adequate to support a highly redundant mesh of links whose characteristics change frequently. The focus of the research is on developing a new generation of wireless data and management protocols that will remove significant barriers to meeting the potential of wireless technologies in the rural environment.

Similar to Australia, Pearson says NZ telcos “have been reluctant to provide broadband access to rural areas because of the low population density and high cost of installation. Despite this, the current government have identified providing broadband access to rural communities as an important factor to economic success of the country into the future. They are currently funding the PROBE (Provincial Broadband Extension) project to encourage the deployment of broad band networks in rural areas.” Wireless technologies are likely to be one of the key technologies used in this deployment, he says.

Waikato is building CRCnet in a rural area to the west of Hamilton. The first link, 13.4km in length at a nominal bandwidth of 11Mbps, has been operational since December 2001. Since this time, they have installed a further four links ranging from 2.5km to 17km in length. To mirror realistic environments, one of the sites is a solar powered repeater half way up a mountain with no power. This solar power site has been operational since March of this year.

“The next three links are in the process of being deployed,” Pearson says. “Experience with this deployment has identified a number of issues, such as poor link efficiency, that must be addressed before large broadband wireless networks can be deployed economically in rural areas.”

Seed funding was provided by the University Department and they have just secured funding of NZ$927 000 (approx. AU$800 000) over the next three years from the New Zealand Foundation for Research, Science and Technology.

“We are currently using Orinoco radio cards running the standard 802.11b protocol,” says Pearson. “These cards are mostly hosted in Linux machines although we are using a small number of access points at present. Most of the Linux machines are three to five year computers that have been retired from their original use.”

He adds, “in places where power is an issue, we have a number of biscuit computers (3.5 inches per side, completely solid state, a single 5V power input and low power consumption) have been used. At present it is marginal whether an installation is cost effective. But as the price of wireless gear continues fall and we develop our own technology we believe we will be able to make it cost-effective.” As the costs of wireless products continue to drop, and the group develop their own technology, Pearson believes the network will be cost effective.

But it hasn’t been all smooth sailing and Waikato has experienced a number of challenges setting up the project.

Firstly, says Pearson, “the wet summer made access to some sites extremely difficult and progress was much slower than expected.”

Sorting through the large wealth of information, some of it conflicting, and deciding what was technically possible and legal was the second key challenge.

He points to power as another key challenge. “Getting power to some sites is either not an option or is very expensive. Solar powered sites are expensive so would need to support a large number of users to make them worthwhile. We are look at a number of lower powered options that make it possible to reduce the cost of a solar powered site.”

Finally, Pearson concludes, “planning the locations for repeater sites has [also] proved much more difficult than expected,” he continues, “relying to heavily on topographical map, which don’t show things like the height of trees, in this process can lead to some surprises. At this stage we are unaware of any substitute for spending lots of time driving around with a pair of binoculars to make sure ‘line of sight is possible’.”

That issue also reflects one of the limitations of wireless in regional areas, according to Pearson.

“The limitations tend to be determined by the landscape,” he says. “For example in densely populated areas the number of channels can potentially be a real problem particularly if multiple wireless ISP’s are operating in the area. In this situation interference can be a real problem.”

He continues, “in gentle rolling country side with lots of trees coverage can be a real problem. While the distances between nodes can be quite small it might not be possible to set up a link between them – one or more intermediate links may be required. This is the ‘line of sight requirement’.”

Despite the challenges, network use has been positive.

“The project started about eight months ago,” Pearson explains, “and we have spent quite a lot of this time learning about the technology and how to deploy it. Nevertheless, at present we have two houses and two schools connected to the network. In the near future we have plans to add another six schools and a reasonable number of houses. Most of the traffic is currently web traffic.”

The group have also held a number successful of video conferences (at 384 kbps) over the network. They are currently looking at a number of new applications that can be used over the network.