Two average weighted measurement fusion Kalman filtering algorithms in sensor networks

For Kalman filter-based data fusion in sensor networks, based on the weighted least squares (WLS) method, two distributed measurement fusion Kalman filtering algorithms are presented in terms of the average weighted measurements and the average inverse-covariance matrices, where the second algorithm is equivalent to the micro-Kalman filter (or mu-Kalman filter) derived from the centralized Kalman filter in sensor networks. Using the information filter, it is proved that they are functionally equivalent to the centralized fusion Kalman filtering algorithm, i.e. they give the Kalman estimators which are numerically identical to the centralized Kalman estimators. They not only have the global optimality, and but also can reduce the computational burden. Two numerical simulation examples verify their functional equivalence.