Decentralized detection and tracking of emergent kinetic data for wireless grids of embedded sensors

The paper proposes methods to detect and track emergent kinetic data representing clouds of physical entities, e.g., clouds of polluting gas, or clusters of autonomous agents, like robots or vehicles. Kinetic data are important entities in Cyber-Physical Systems as they express phenomena that are distributed in space and time, and are dynamic with respect to their characteristic attributes and lifetime. The main attributes of kinetic data are topography (position, boundary, and area), nature (signature), and dynamics (up to n-th order gradients). The related operators define the distribution (discretization) and composition (aggregation) of the attributes. The paper presents fully decentralized methods for optimized implementation of kinetic data on a grid network of wireless embedded sensing nodes. The efficiency of the algorithms is compared with similar methods proposed in the literature.