An overview of non-centralized Kalman filters

The usage of wireless sensor networks (WSNs) for state-estimation has recently gained increasing attention due to its cost effectiveness and feasibility. One of the major challenges of state-estimation via WSNs is the distribution of the centralized state-estimator among the nodes in the network. Significant emphasis has been on developing non-centralized state-estimators considering communication, processing-demand and estimation-error. This survey paper presents different methodologies to obtain non-centralized state-estimators and focuses on the estimation algorithms and their implementation. The temperature distribution of a bar is used as a benchmark to assess the non-centralized state-estimators in terms of estimation-error and communication requirements.