Decentralized algorithms for monitoring moving objects with geosensor networks

<p>This research aims to design efficient, decentralized spatial algorithms for monitoring moving objects with wireless geosensor networks. Wireless geosensor networks present new opportunities for monitoring geographic movement, as well as new challenges for spatial computing. In particular, the limited sensing, power, communication, and computing resources of geosensor networks present constraints to traditional, centralized spatial computing data structures, algorithms, and models.</p> <p>This work explores the use of decentralized spatial algorithms for monitoring moving objects via two complementary perspectives: density-based groups of mobile geosensors, and moving objects tracked relative to fixed-position geosensors. A series of decentralized algorithms is described, specified, implemented, and evaluated experimentally via agent-based simulation. The empirical evaluation of algorithm performance is assessed through the accuracy and efficiency of the developed algorithms.</p> <p>The approach and results contribute to answers to the following underlying research questions: Can decentralized spatial algorithms enable coordinate-free monitoring of moving objects via geosensor networks, without centralized control or global data access? Can such decentralized algorithms accurately monitor meaningful movement patterns for moving objects of interest? And what strategies can help decentralized algorithms monitor movement accurately, whilst maximizing communication efficiency and minimizing resource overheads for the geosensor network?</p> <p>The results of this research have relevance to detecting and monitoring salient patterns of movement - such as flocks, convoys, and leadership - amongst moving entities across a range of important application areas. In turn, this work contributes to the development of new capabilities in environmental monitoring and animal behavior; traffic management and smart cities; and emergency management and response.</p>