Tracking of space debris via CPHD and consensus

This paper proposes an algorithm for distributed space debris tracking over a network of sensors with processing and communication capabilities. A dynamic model was used to propagate precise orbital trajectories of space debris. Gaussian Mixture-Cardinality Probability Hypothesis Density (GMCPHD) was used to estimate the state of known and previously unknown space debris. Data incest is a common problem in distributed tracking system. Therefore the consensus algorithm was combined with GM-CPHD filtering to effectively counteract the effects of data incest in a sensor network which has no centralized coordination and with unknown and possibly time-varying topology. The effectiveness of the proposed approach is demonstrated via simulation experiments.