The mesh potato network

A South African project develops a cheap phone system for rural areas

Steve Song

A group of software experts, technical specialists and telecommunications entrepreneurs in South Africa is working to develop an inexpensive system to provide rural and under-served area with affordable telephone communication.

Orange Farm, about 45 km south of Johannesburg, is the largest informal settlement in South Africa. More than 300,000 people live there, but access to water and a reliable electricity supply is difficult for many of them. There are a few internet cafés and there is mobile phone coverage, but the cost of these services makes them inaccessible to most of the residents. One company, however, has developed a local wireless network to make calls much cheaper and is developing the system to bring affordable telephone services to rural areas.

Dabba, a South African telecommunications company, initially based its system in Orange Farm around an existing internet connection in a local community centre. The company installed a number of wireless routers that broadcast the internet signal over a larger area. Anyone within range of the network can connect and call using either a Wi-Fi or SIP (session initiation protocol) telephone, which look similar to mobile phones but are designed to connect to wireless networks.

Calls made within the network are free, while customers who want to call mobile or fixed telephone numbers can buy pre-paid vouchers from Dabba. The cost of these calls is lower than those normally charged by the mobile phone operators or Telkom, South Africa’s largest telecommunications company.

Although the system provides cheaper phone calls to the people of Orange Farm, it is still relatively expensive to buy a telephone capable of connecting to the network (prices of Wi-Fi or SIP phones average at around US$100). The investment needed to build up such a network is also considerable and requires specialist technical skills. This combination of factors means that installing a similar system would be beyond the reach of many communities.

Smart solutions

Dabba’s basic business model attracted the attention of the Shuttleworth Foundation, a non-profit charitable organization, also based in South Africa. Working together with Dabba, and a broad partnership of international experts, the Foundation set up the Village Telco project to develop a telephone communication system that would be affordable for rural and other communities with limited telecommunications services, and which would require little technical knowledge to install.

The project team decided that the whole system should cost no more than US$5000 to get started, an amount that a small-scale entrepreneur, someone who already owned an internet café, for example, would be able to afford. To minimize the financial risk, the team also had to come up with a business plan where investors could expect to break even within six months. Using this model, rural villages, and other areas with few alternative communications options, would have the opportunity to develop their own telecommunications company, or Village Telco.

Building a wireless network with a range of a few kilometres is fairly straightforward. The biggest problem for the project team was to overcome the need for Wi-Fi or SIP telephones. These telephones are expensive and usually have a very limited range, working only within 100 metres of an internet signal. This would restrict the coverage of the network and the number of people who could afford to buy a phone.

Extending the network, however, would mean adding more access points. This could be done by using wireless routers, similar to those used in a home wireless network, but a large number of routers would mean more expense. The cost would quickly rise above the project’s limit of US$5000. The team was left with two options: find some way to extend the reach of the phones while somehow also making them cheaper, or find a way to reduce the cost of adding the access points.

The latter option seemed the most likely, and mini-routers are already available for a reasonable price. As long as the customers were within range of one of these small routers they could connect to the network. But they would still need an expensive Wi-Fi or SIP phone.

Traditional telephones, also called POTS (plain old telephone service) phones, are much cheaper but they cannot connect to a wireless network without an ATA (analogue telephone adaptor). This device converts the analogue signal from a standard phone into the digital signal needed to connect to the internet. But having these two extra pieces of equipment, the mini-router and the ATA, would only complicate what was supposed to be a simple system, and would add to the cost. What the team needed was an inexpensive router that could also convert a standard analogue telephone signal to digital. What they came up with was the ‘Mesh Potato’.

Package deal

The type of wireless network used in Dabba’s system, where the signal travels between several routers, or nodes, on its way to and from the original internet connection, is known as a mesh network. The Mesh Potato basically couples a wireless router with an ATA. It can convert the analogue signal from a standard telephone then connect to the network to transmit the new digital signal over the mesh network.

To keep costs low, the Mesh Potato runs on open source software, and has been developed so that it is easy to install and uses very little energy. All the necessary components are contained within a casing strong enough to be kept outdoors, if necessary, at a total cost of no more than US$60 per unit.

The idea is that the Mesh Potato would be part of a subscription package, an initial one-off payment that would enable the customer to connect to the network and start making calls. The Mesh Potato would come with its own pre-programmed phone number, supplied by the Village Telco operator, in much the same way as a mobile operator sells mobile phones. The customer can then fit the Mesh Potato to the side of their building, or inside their home, and plug in a standard telephone.

But this is only part of the story. The internet signal still has to get to the Mesh Potato. The team quickly realized that in order to get started with the maximum impact, the system would benefit from a single ‘super node’. This is the first external access point where the wireless signal is taken from the original internet source, from an internet café, for example, and broadcast over a wider area – to the Mesh Potato in the customer’s business or home.

One way to achieve this is to have special wireless routers, or access points, that can transmit the internet signal over a large distance and mount them on an antenna. Three such routers, pointing in different directions, would be able to transmit the signal for a radius of about 1 km, after which the signal is picked up by one Mesh Potato which, in turn, passes it to another Mesh Potato, then to the next, and so on. The more Mesh Potatoes there are in the system, the wider the coverage.

When put together, these pieces of equipment connect and form the mesh network that carries the internet signal as follows:

  • The internet signal enters a computer server in the Village Telco office (perhaps an existing internet café) via a broadband cable connection or satellite link.
  • The signal is then carried to an antenna on the roof of the Village Telco premises, or other high place.
  • The signal is then transmitted by the ‘super node’– three external access points placed around the antenna providing 360° coverage.
  • The Mesh Potato, placed in the customer’s office or home, picks up the signal, which is carried throughout the network from one Mesh Potato to another.
  • The customer plugs a standard telephone into the Mesh Potato and makes a call.
  • The Mesh Potato converts the telephone signal from analogue to digital and transmits it back to the super node, or to another Mesh Potato on the network, which will keep passing the signal on until it reaches the super node.
  • The signal then reaches the antenna and arrives back at the server computer in the Village Telco, connecting the caller with the internet and other telephone networks.

What the project team have done is to think through all the problems that could arise in building a small telecommunications network. The intention is to solve most of the technical difficulties and produce and system that can be easily installed and maintained without expert technical knowledge. Similarly, customers should be able to take home their Mesh Potato, switch it on and start calling.

Longer links

The biggest challenge of the Village Telco, however, is to connect the network to the rest of the world. Doing so can be a good source of income for the company as other telecommunications companies pay a dividend for every connection made to their network. Dabba have invested in the technology to connect to the mobile and fixed-line operators in the rest of South Africa. But this process is expensive, time consuming and requires some expertise. Not all Village Telco operators would be able to set up these connections

The idea is that companies like Dabba would provide the wider network to the smaller Village Telcos, enabling them to connect to mobile and fixed-line services. Individual village operators therefore don’t have to negotiate with multinational firms or invest in extra equipment, they just connect to the network already installed by larger telecommunications companies.

Some of the smaller Village Telco operators will have the time, skills and money to make their own connections with the larger providers. If they develop their own network, perhaps covering a particular region, then they could compete with Dabba and help to keep customer prices low. Having multiple operators also gives the smaller Village Telcos a choice of who to connect with, which may not always be based only on price but on who delivers the best service.

But the Village Telco system is being developed to give people living in rural areas the chance to communicate with each other, even over relatively large distances. And a connection to the wider, global internet is not necessary to provide voice services to a community. The network can operate independently, meaning that anyone living within range of the wireless signal can call anyone else within the same network. Customers just have to buy the Mesh Potato and a cheap telephone.

In rural areas, where the population is less concentrated, the business model would probably have to be slightly different, with perhaps one organization or institution carrying most of the initial cost. It is unlikely that they would see a return on their investment in such a short time as six months but they would be providing a useful service to the community.

Adaptable

The Village Telco project team is deliberately designing the system to be flexible so that, in future, telco operators will be able to adapt the software to provide financial services, similar to the banking services provided by some mobile operators. Customers would be able to transfer credit from the pre-paid vouchers from one person to another to buy goods or contribute to a savings scheme.

The software, Asterisk, can also be customized to develop an interactive voice response system. A farmer could call a number on the network, for example, which could even be a free call, and follow recorded instructions. Pressing 1 on the telephone could give a local weather forecast, number 2 might give market information, and 3 could provide pest control advice. These services could be easily updated according to local needs and conditions.

In the meantime, Dabba services in Orange Farm continue to grow. Members of the project team have had some discussions with entrepreneurs in Khayelitsha, another informal settlement near Cape Town, who are interested in the system. There is also some interest from manufacturers in China to produce the Mesh Potato and the project developers hope to have the software package ready to download by the end of 2008.

The Village Telco team are quick to point out that while the network might be innovative, they are not inventing anything new. The technology already exists, it just hasn’t been put together in quite this way before. Even the Mesh Potato is an amalgamation of software and hardware that is readily available. And this is one major advantage of the system; the technology has already been tried and tested, the team are simply putting it all together to make it easy, affordable and flexible. It is now up to the communities and villages who will use the technology, to adapt and create a system that perfectly suits their needs.

----------------------------------------

Steve Song is a telecommunications fellow with the Shuttleworth Foundation.

Village Telco
The Village Telco website with regular updates on the project’s development.

Many Possibilities
Steve Song’s personal blog on telecommunications and technology for development.

Dabba Telecom
Dabba Telecom develops community-owned affordable voice and data services for areas with little or no access to communication services.

Inveneo
Inveneo specializes in developing communications technology for rural areas of developing countries.

Free Telephony Project
The project provides free and open source hardware and software designs for telephone systems.

Council for Scientific and Industrial Research
CSIR is a South African scientific and technology research organization specializing in research and development for socio-economic growth.

30 September 2008

Copyright © 2014, CTA. Technical Centre for Agricultural and Rural Cooperation (ACP-EU)