Evaluation of Kalman Filter-Aided GIPS for Passive ToF 3D Sensing

Passive 3D source localization and sensing have drawn considerable academic attention, propelled by their broad applicability in wireless communications and sensing. We aim to ascertain the dynamic trajectories of the source, building upon earlier studies that focused on a linear approach. The paper acknowledges the limitations of linear trajectories in capturing complex behaviors in ellipsoidal-constrained passive imaging. Therefore, this work extends the performance evaluation for a wider range of trajectories. No work in 3D passive imaging has studied whether these trajectories are meaningful enough to be exploited effectively in source localization. Our analysis showed that the random walk performed better in the circumstances we studied, showing considerable deviations from standard linear trajectories. This work not only validates the current paradigm, but also presents novel opportunities for practical implementation in passive imaging for source localization. Different trajectory models are generated in simulations. We use simulated dataset to evaluate algorithmic performance and showcase the effectiveness of considered emitter motion models for joint 3D source localization and passive imaging.