A Federated Kalman Filter Design Using a Gain Fusion Algorithm

Abstract

As an alternative fusion process of the federated Kalman filter, a gain fusion algorithm is newly proposed in this paper. In this algorithm the optimal covariance and estimate are obtained by using local Kalman gains and estimates. Consequently, this algorithm reduces the amount of communications and avoids the need to calculate inverse covariance matrices in local filters. It is mathematically shown that the suggested algorithm guarantees the global optimality when all local sensors produce equivalent information except their precision. It is prospected that this algorithm may be well suited for implementation of the multisensor navigation systems.