Social Sciences · Sociology and Political Science
Information Quality Aware Sensor Network Services
Dinkar Mylaraswamy (honeywell, Mani Srivastava (ucla
See on the mapAbstract
Abstract: Wireless sensor networks have proven useful for applications in diverse domains. The challenges of scale and resource constraints posed by these systems have led to development of novel network protocols and services, but their focus has been on traditional metrics of quality of service of network data transport. Rather, sensor networks require combining networking quality of service concerns with metrics of quality and integrity of sensor data sources and performance of sensor fusion algorithms. We describe how network protocols, data integrity management, and sensor fusion algorithms can be design to cooperatively optimize the "Quality of Information" returned by a sensor network. Extended Summary High-fidelity and real-time observations of the physical world are critical for many applications in military, scientific, medical, industrial, urban, social, and personal settings. With their ability to sample physical world processes in spatially and temporally dense fashion, wireless sensor networks are able to make these observations with unprecedented details and perspectives. In a typical wireless sensor network, measurements from distributed sensors are aggregated to reconstruct spatiotemporal behavior of desired physical variables or to detect, identify, and localize sources and events of interest. The system is designed under constraints on cost, bandwidth and energy resources while optimizing performance