Wireless sensor networks that have the capabilities of sensing, computation and communication make an ideal solution for emergency warning systems. Appropriate authorities expect immediate warnings predicted by such systems. This research designs, implements, and evaluates an energy-efficient and real-time data delivery model which can be employed in wireless-sensor-network-based emergency warning systems. With the new data delivery model being used in an emergency warning system, warnings can be obtained in real time, and network lifetime can be extended.
Different from the traditional 'data collection' and 'data aggregation' methods, this research focuses on the stringent requirement of delivering warnings in real-time. Instead of managing a large amount of raw data, a simpler mechanism for event detection and reporting that can reduce energy consumption as well as notification delay is investigated. In addition to the impact on practice, the theoretical analysis on the bounds of energy consumption and notification delay has a strong impact on the theory aspect. Moreover, this project is a starting point to integrate research and education for the purpose of attracting both undergraduate and graduate students to the area of wireless sensor networks.
