Adaptive data collection in sensor networks

Overcoming the disparity between the amount of data a wireless sensor network can transfer before depleting its limited energy resources is an ongoing challenging. This is particularly important when the entire state of the network is required, such as when a global predicate must be evaluated. This paper presents an scalable adaptive data collection algorithm for gathering all sensor node readings. A novel paradigm, where observed sensor node readings and locations are encoded as bitmap data that conserves the spatial relationship between sensor nodes, is utilized. The technique provides a large reduction in the volume of data transmitted in a both dense and sparse sensor networks when attempting to gather an initial global snapshot of the system.