Identity Management of Clustered Satellites with a Generalized Labeled Multi-Bernoulli Filter

Tracking clustered satellites in the presence of noisy measurements creates a scenario where satellites are likely to be cross-tagged. This work focuses on the identity management of a closely spaced set of CubeSats immediately after a deployment. The main contribution is an identity management process that augments the generalized labeled multi-Bernoulli filter to probabilistically handle identity assignments. The effectiveness of the method is demonstrated on simulated tracking data from the 88 Planet Labs Flock 3 CubeSats deployed on PSLV-C37. Simulation results indicate that the proposed identity management process can effectively establish identities within 3 days after deployment.