Abstract
Smart sensors possessing wireless communication and computational capabilities are expected to allow dense measurements of structures and to elucidate their performance. While functionalities such as synchronous sensing, reliable data transfer, and efficient data aggregation need to be improved, available power, sensing, and computational resources are limited. In view of such limitations, middleware services, i.e., 1) resampling-based synchronized sensing without strict timing control, 2) reliable data transfer with low acknowledgement message requirements, and 3) collaborative and distributed data aggregation are proposed. The services are implemented on the Imote2 smart sensors and experimentally validated.
