ABSTRACT
This chapter discusses different strategies to identify and treat dependencies among Kalman filter estimates while pointing out advantages and challenges. It reveals dependencies between locally processed data as a major challenge for networked state estimation. The chapter then focuses on multisensor filtering principles. State estimation methods are utilized to provide insights into a system's behavior. An estimate for the system's state is dynamically computed based on prior information, a process model, and measurements stemming from sensor devices. The Kalman filter algorithm embodies an optimal solution to the state estimation problem providing estimates that minimize the mean-squared estimation error. The distributed Kalman filter can be characterized as a further development of the federated Kalman filter and provides the optimal estimation result after any number of time steps whereas the federated formulation is optimal only in special cases. The Bar-Shalom-Campo formulas represent the optimal solution to the fusion problem and are well-known in multisensor tracking applications.
