We present coSense —the collaborative, fault-tolerant, and adaptive sensing middleware for the Internet-of-Things (IoT). By actively harnessing the greatest asset of the IoT, the sheer number of devices, coSense is able to provide easy data acquisition with quality-of-service-based data cleaning by combining unsupervised learning and information fusion. It can also greatly improve sensor accuracy and fault tolerance to produce measurements specifically tailored for modern data-driven IoT empowered applications. In this article, we focus on the general concepts behind coSense and evaluate the accuracy gain based on a real-world dataset.
CITATION STYLE
Schmeißer, S., & Schiele, G. (2020). coSense. ACM/IMS Transactions on Data Science, 1(4), 1–21. https://doi.org/10.1145/3395233
Mendeley helps you to discover research relevant for your work.