If you’ve traveled abroad in remote locations – or driven across the Upper Peninsula – you know that internet connections in rural environments can be unstable or nonexistent.
Facebook and the University of Michigan aim to fix that. In a collaboration with U-M, the Universidad Politécnica de Madrid and the Ohio State University, Facebook is funding an open source initiative designed to improve low resource setting connectivity. Their Connectivity initiative is pursuing a variety of programs like this to develop advances to make connectivity more accessible and equitable around the world. One key area of research is in wireless communications, which can provide low-cost long-range broadband internet access.
Seeing through mountains
Line of sight is often a necessary component for broadband, long-range wireless signal transmission from a high resource area to a remote location. In many parts of the world, things like terrain, foliage and weather interfere with the ability of a signal to get from one place to another.
Facebook Connectivity wanted to find ways around that problem. “We turned to the experts,” says Julius Kusuma, Research Scientist at Facebook Connectivity Lab. “We knew we needed a coalition of academics working together on the concepts with us to do things we couldn’t do on our own. Connectivity challenges are complex and multi-dimensional, and by working with academic and industry partners with expertise in wireless communication, radio propagation modeling, electronics, algorithms, econometrics, and more, we could find solutions together.”
The technology backbone
Rugged and hilly terrain, foliage, and throwing in cold and snow or heat and humidity, often drive up the cost and complexity of connectivity. These barriers are often prohibitive in rural or less resourced locales. The backhaul, the backbone of internet connections, in rural cases is often a wireless point-to-point link. The Facebook consortium seeks ways to enable network designers to use this phenomena where it is possible to provide reliable connectivity.
“Today, the practice in network design dictates that electromagnetic waves need direct paths. We want to augment this practice for cases where there is no clear line of sight, to avoid the need to build repeaters. We need to develop signal modeling tools that can reliably predict signal strength through diffraction. . With advances in computation (U-M’s domain expertise), we are developing better modeling techniques that can enhance predictability of how much connectivity there will be when there is no line of sight,” says Eric Michielssen, Associate Dean for Research at the College of Engineering and U-M lead in the consortium.
Connecting the consortium
The consortium, pre-pandemic, got the chance to gather for a workshop to meet with industry experts, present findings and develop the next phase of the research. The consortium is testing the technology in specific locales that offer a variety of environmental challenges. Those real-world tests will help the consortium iterate until models reach an acceptable level of predictability. Facebook has published a playbook on rural connectivity and a white paper on the project.
Says Michielssen, “Julius – and Facebook – have been great to work with. It’s exciting to be connected to a company and group so forward-looking and that moves so fast. This project has opened up new avenues for collaboration and forced us to look at some of these problems in new ways.”
Kusuma concurs, “Our academic partners brought data and measurement we couldn’t have done ourselves and done it more rapidly than we expected. This has been a mutually beneficial collaboration.”