Utility-Driven Spatiotemporal Sampling Using Mobile Sensors

Many real-world applications for sensor networks require event sampling with sufficient resolution over both spatial and temporal dimensions. When the deployed nodes are insufficient to fully cover the sensor field to satisfy this spatiotemporal sampling requirements of all events, nodes with intelligent mobility are crucial to improve the sampling quality. We measure the sampling quality using a utility function that incorporates both the importance and spatiotemporal properties of events. We then describe a Utility Driven Mobility (UDM) scheme, which enables autonomous mobility scheduling of nodes in a distributed fashion. We evaluate the performance of UDM via extensive simulation studies. Our work provides a practical framework for spatiotemporal sampling using mobile sensor networks.