The term "wireless sensor network" (WSN) describes a system that comprises simple nodes capable of sensing, computing, and communication. Conventionally, such WSNs are intended for rather simple applications and are intended to operate for long duration on battery power only.
Recently, however, the areas of applicability has broadened. Systems like Amazon Echo have shown that acoustic sensing has great market potential; teleconfereing systems have used distributed microphone arrays in various settings, yet usually based on wired networks. The scenario of multiple microphones, connected by wireless communication has been largely unexplored and is the topic of a DFG-funded research project on Acoustic Sensor Networks (ASN).
This ASN project provides the context for this project group. We will develop a testbed in which acoustic signal processing algorithms can be deployed onto real hardware (Raspberry Pi 3) and experiments can be run under controlled circumstances. We will develop mechanisms to deploy such algorithms onto the testbed and how to chain together multiple stages of a signal processing pipeline across multiple nodes (possibly using REST APIs). At the end, a working prototype of such a testbed should be available.
This project is jointly organized by the Computer Networks group (Prof. Karl) and the Communications Engineering group (Prof. Häb-Umbach).
Slides about the project group.