Distributed Three-Dimensional Pseudo Measurement Kalman Filter with Angle-Only Measurements

This paper addresses the problem of three-dimensional target localization using a network of sensors with bearing-only measurements. We derive a distributed pseudo measurement Kalman filter by utilizing the concept of information weighted consensus. Theoretical analysis reveals that the estimation output is biased due to the correlation between the pseudo measurement matrix and the pseudo measurement noise. We then propose a distributed instantaneous bias compensation algorithm to mitigate this issue. Extensive numerical simulations reveal that the proposed approach can approximately recover the performance of the centralized solution and outperforms other mainstream distributed algorithms.