Efficient time propagation of U-D covariance factors

Time propagation of the Kalman filter covariance matrix involves an operation of the form . For many applications has a block triangular structure. When the filter implementation employs covariance factorization (i.e., recursions for and are used, where , with unit upper triangular and diagonal) this matrix structure can be used to reduce the computation cost of time propagation. The contribution of this paper is a streamlined Gram-Schmidt orthogonalization algorithm that can dramatically reduce time update computation costs.