The goal of the WaveScope project is to build a sensor computing platform for high data-rate, disconnected sensor network applications. It is motivated by industrial, structural, and automotive monitoring applications that use many embedded, wireless, and high-rate vibration, pressure, magnetic, and sound sensors to detect faults, anomalies, and trends. Timely detection and diagnosis can reduce maintenance costs and assist with troubleshooting. The project employs three classes of methods to facilitate the construction of such applications:
1. It provides a general-purpose information processing architecture that combines continuous query processing and signal processing on sensor data streams. 2. It employs in-network processing algorithms based on probabilistic models and wavelets that adapt to prevailing network conditions. 3. It uses a multi-hop wireless data delivery system that provides efficient and fair network resource allocation in the face of intermittent connections via diverse links (e.g., 802.15.4, 802.11, BlueTooth, and others)
The WaveScope infrastructure allows engineers and scientists to rapidly develop high data rate sensor-based applications. Target WaveScope-based systems include automotive sensor networks and industrial equipment monitoring.
In addition to research papers and impact on courses, expected results include: High data-rate protocols and information processing architecture. Open-source software that runs on commercial hardware and allows others to quickly develop and deploy high data-rate sensing applications. A real-world deployment and evaluation in one of the target application areas.
