The challenges of the technical implementation of the wireless mesh network at Zimpeto Market

Nuno Santos
Aug 30, 2018 · 5 min read

The Wireless Mesh Network Project, which has already been reported on by Suzana Moreira, can be described as a low cost, high gain, wireless network solution designed for low income target audiences who have little to no access to digital information. For a more detailed reminder of what a mesh network is, read Bojan’s post describing the technology in simple terms.

This post comes from the perspective of Champier Lda., a tech start-up who, in partnership with Mowoza, are tasked with the technical implementation of the Mesh Network in the field, and who have been involved in the scoping for the project from the very beginning.

There are, however, many challenges that, no matter how much effort is put into scoping, theoretical studies or even physical surveys of the site, cannot be predicted when it comes to the implementation of a project such as this and in sites such as the one picked for this specific project: the Zimpeto Market, in the outskirts of the Mozambican capital of Maputo.

With the goal of conducting a preliminary, small scale, test of the network (resonant with the ‘start small’ ethos of the Frontier Technology Livestreaming), Champier Lda. headed to Zimpeto Market to meet with a Mowoza collaborator and measure some important data, needed for assessing the needs for full scale implementation.

The first challenge, before any technical work could be done, was regarding the security of the equipment necessary to implement the network. Zimpeto Market is an outdoor, public space, where any electronic device installed runs the serious risk of theft, within a very short period of time.

This meant that even for the testing phase, no equipment could be left unattended. For the purpose of the activity at hand, however, there was authorization to use the Municipal Office premises to fix the device which would be used for the measurements.

It is important to clarify that, for this project, a security solution was scoped for the housing of the devices that would transmit the signal in the outdoor areas, but there would always be a need for an indoor space to store a device that is central to the operation of the Mesh Network, which is the server. More on that later…

There are many cement buildings adjacent to the central market court where the devices and metal security housings could be fixed, with the use of metal brackets screwed into the walls.

What we wanted to ascertain, first and foremost, was whether the electrical grid in place at the market could support the network infrastructure. It was found that, since the basic costs of the market are supported by the municipal authorities, the electrical grid is open for all the vendors to connect to. It would be a simple matter of connecting wires from the electrical grid to the devices that need it. The devices do not consume a significant amount of power, so there would be no risk of overload even if that was not covered by the authorization that was already granted for the project, by the Municipal Authorities.

Layout of the market, with the central open space, and stores built around it.
The electric infrastructure proved adequate for any kind of network implementation proposed in the market, as there are many points in the grid from which and electric cable could be extended and connected to network transmitting devices.

Secondly, we needed to find out if the device chosen to transmit the network signal would perform appropriately in the specific environment of the market, specifically regarding the range and quality of the signal. This would allow us to decide how many devices would be used to cover the whole market, for the full scale implementation. The findings on this point were satisfactory, firstly, due to the layout and size of the market, which is made up of a central open space surrounded by solid structures and secondly, due to the fact that those structures have walls that are not thick to the point of hindering the transmission of the signal.

Reach of the network signal in the middle of the market.

After the measurements it was easy enough to calculate that two or three devices would suffice to cover the whole market with the network.

Scheme of the signal propagation with the full scale network in place.

Upon conclusion of this testing activity, the next steps would be, on the one hand, to project the costs and other needs for full implementation, in terms of hardware acquisition and assembly and, on the other to begin scoping the software component of the Mesh Network Solution.

A lot more can be said about the scoping for the software development, but a significant amount of work is still to be done on this front, as it takes another set of steps to test and conceive the best ways to convey content to the public, through the Mesh Network.

From a technical perspective, and in conclusion, the only question left in the open is related to the security issues addressed at the very beginning: there needs to be an indoor space to house the server that will be the management hub of the network. The solution could be in the hands of one of the owners of the stores that are built around the open space outdoor market, who could possibly allow the server to be placed in their store. Therein lies the challenge that has always been at the heart of the project: the gauging of the relevance of the project, as perceived by the people who will be the end users. In that sense, convincing a small sample of the target audience that the Wireless Mesh Network could improve their lives could more than ever be the key for the success of this project.

Follow this Medium page and stay tuned for more updates!

Frontier Technologies Hub

Working with the FCDO to apply frontier technologies to development