A state-space approach to QRD-LSL interpolation and QRD-LSL smoothing

Sayed and Kailath (1994) demonstrated the feasibility of directly deriving many known adaptive filtering algorithms in square-root forms by a proper reformulation of the original adaptive problem into a state-space form. This work employs this state-space form to develop adaptive interpolation and smoothing algorithms. In particular, a systematic and concise derivation of the QR-decomposition least-squares lattice (QRD-LSL) interpolation and smoothing algorithms using correspondences between Kalman filtering and LSL adaptive filtering is given.