Hybrid, Large-Scale Wireless Sensor Network for Real-Time Acquisition and Tracking

摘要

This thesis proposes a hybrid, large-scale wireless sensor network (WSN) designed to support real-time target detection and tracking of multiple ballistic missile threats. In particular, the proposed WSN consists of terrestrial as well as satellite nodes. The IR signatures presented by the target-background combination are explored and modern IR sensor technologies are examined in search of a suitable IR sensor for the proposed hybrid, large- scale WSN. A multicolor, Quantum-Well Infrared Photodetector (QWIP), step-stare, large-format Focal Plane Array (FPA) is proposed and evaluated through performance analysis. The thesis proposes an efficient data dissemination mechanism as well as a suitable medium access control (MAC) scheme for the proposed WSN that is designed to meet real-time and accuracy requirements without introducing excessive overhead and increased end-to-end time-delays. A clustering mechanism, called the 'Area of Interest' (AOI) is introduced, which combines the content-based feature of the data centric routing approach with the principles of in-network data aggregation and clustering. Simulation results verify that aggregation within the AOI improves the data throughput across the full range of network load. A contention-based MAC scheme, Carrier Sense Multiple Access (CSMA), and a contention-free approach, Time Division Multiple Access (TDMA), are examined. Performance analysis and simulation results indicate that a contention-free approach is suitable for implementation in wireless networks associated with large propagation delays and increased offered loads. Matlab and OPNET Modeler (copyrighted) software packages are used to simulate and evaluate the proposed schemes.