Discrete-time fixed-lag smoothing algorithms
Automatica (Journal of IFAC)
Fixed-lag demodulation of discrete noisy measurements of FM signals
Automatica (Journal of IFAC)
1972 IFAC congress paper: Optimal minimal-order observers for discrete-time systems-A unified theory
Automatica (Journal of IFAC)
Discrete-time demodulation of continuous-time signals
IEEE Transactions on Information Theory
A view of three decades of linear filtering theory
IEEE Transactions on Information Theory
| Hi-index | 22.14 |
An alternative to continuous-time Kalman filtering is sampled digital processing. The purpose is to develop a good model for this situation and investigate the resulting performance. The error dependence on the sample rate is studied. It is found that a sample rate considerably higher than twice the half power-frequency is necessary to avoid loss of performance where zero-lag filtering is desired. Smoothing is discussed and in this context three results of major importance are found; (1) an optimal sample rate exists for each smoother (2) only a few steps of lag is necessary to approach the ideal unrealizable error (3) a considerable lower sample rate can be accepted for smoothing as compared with filtering. This finding is of interest as the corresponding ideal continuous-time smoother has been reported unstable. Finally a comparison is made with approximate models suggested in the relevant literature.