Target localization: Energy-information trade-offs using mobile sensor networks

We investigate the energy-information trade-offs in target localization using a group of N sensing agents based on bearing measurements. We aim to design optimal trajectories for the coordinating sensing agents that minimize the kinetic energy and maximize the information collected by the agents along their trajectories. The determinant of Fisher information matrix (FIM) is used as a metric for the information. Inspired by biology, we construct a constrained calculus of variations problem that captures the observation that better information results in more energy dissipation, which requires higher energy supply. We solve the equations of motion as Hamilton equations, which produce a set of ordinary differential equations that can be integrated for the trajectories of the agents. An extended Kalman filter (EKF) is used to produce estimates of the state of the target.