Extension of DGKCF to multi-target tracking

Extend the Dynamic Generalized Kalman Consensus Filter framework to multi-target tracking applications in sensor networks with switching communication topologies.

Background

The paper evaluates DGKCF for tracking a single target using a linear time-invariant state and measurement model, with sensor agents communicating over switching network topologies and some agents being oblivious. Its consensus weights are computed from local prior covariance information.

The applicability of this framework to scenarios involving multiple simultaneously tracked targets is left unresolved. Extending DGKCF to multi-target tracking would require adapting its distributed information fusion and consensus mechanisms to handle multiple target states and the associated sensing and communication requirements.