Suboptimal approaches to the one-step fixed-lag smoothing problem for
Markovian switching systems are examined in this paper. Two algorithms
are developed for obtaining one-step fixed-lag smoothed estimates. Th
e first algorithm is an approximation that considers the models over t
he two most recent sampling periods, while the second one considers on
ly the models over the most recent sampling period. Simulation results
are presented to compare the performances of these smoothing algorith
ms.