Doing large-scale computations on an Internet of Things network

As you may notice, the Internet-of-Things devices are everywhere: in our factories, in our cities and in our houses. The challenge of tomorrow will be to store and process data generated by these devices. In this master thesis, we tried to tackle this challenge. We implemented a MapReduce algorithm designed to run on the extreme edge of the network. With the edge computing, the IoT devices can exchange, store and process data without depending on an external infrastructure. As part of this master thesis, we developed a resilient MapReduce algorithm that work in a peer-to-peer network composed of IoT devices.