Marine Vehicle Tracking using Agile Maritime IoTs

Abstract

Marine surveillance deals with the complex problem of modeling and estimating natural and human processes while operating with limited information acquisition resources. The complexity of the problem and the scarcity of resources, such as power, processing, and communications, requires agile distributed sensing and computing that provides scalability, diversity in information acquisition, and adaptive allocation of limited resources. This work deals with tracking multiple marine vehicles using data from heterogeneous sensors. Accurate tracking is achieved by a synergy between a network of agile maritime IoTs, with limited processing resources, and a fusion center, with computationally intensive capability. The IoTs can reconfigure their elementary processing operations settings to produce informative yet low volume measurement statistics. The fusion center handles heterogeneous data fusion and solves the complex problem of vehicle state estimation. We demonstrate that by configuring its agile distributed computing capabilities, the proposed system provides significant savings in processing and communications resources without deteriorating tracking performance.