Wireless smart sensor networks have become an attractive alternative to traditional wired sensorsystems in order to reduce implementation costs of structural health monitoring systems. Theonboard sensing, computation, and communication capabilities of smart wireless sensors havebeen successfully leveraged in numerous monitoring applications. However, the current data acquisitionschemes, which completely acquire data remotely prior to processing, limit the applicationsof wireless smart sensors (e.g., for real-time visualization of the structural response). Whilereal-time data acquisition strategies have been explored, challenges of implementing highthroughputreal-time data acquisition over larger network sizes still remain due to operating systemlimitations, tight timing requirements, sharing of transmission bandwidth and unreliablewireless radio communication. This report presents the implementation of real-time wireless dataacquisition on the Imote2 platform. The challenges presented by hardware and software limitationsare addressed in the application design. The framework is then expanded for highthroughputapplications that necessitate larger networks sizes with higher sampling rates. Twoapproaches are implemented and evaluated based on network size, associated sampling rate, anddata delivery reliability. Ultimately, the communication and processing protocol allows for nearreal-time sensing of 108 channels across 27 nodes with minimal data loss.
展开▼
