The primary goal of the project is to develop the architectures and the algorithms needed to exploit mobility of nodes to enhance the performance of sensor networks. It is based on the premise that mobility can be a major asset for both information exchange and collaborative decision-making in sensor networks. The project envisions a sensor network with four different types of node mobility: (i) no mobility, (ii) uncoordinated mobility, (iii) pre-planned mobility, and (iv) adaptive mobility. Each type of mobility offers a different set of opportunities and challenges with respect to optimizing the bandwidth and energy consumption needed to meet the accuracy requirements of collaborative decision-making. The project will develop adaptive sensing and the associated information exchange strategies to carry out this optimization. The expected results from the project will include: (i) algorithms for a base set of communication patterns needed to efficiently support information exchange in sensor networks, (ii) adaptive algorithms for path planning, data sampling, and information fusion for selected collaborative decision-making problems such as source localization and field estimation, (iii) characterization of latency, energy, bandwidth, and decision accuracy trade-offs in the new algorithms, and (iv) demonstration of the effectiveness of the new algorithms through simulation studies and experimentation with data from a testbed. The expected results have the potential to impact many areas of society including intelligent transport systems, homeland security, medicine, public safety, and environment. Dissemination of the results will occur through publications in conferences and journals and through undergraduate and graduate courses at University of Wisconsin, Madison.
