Digital Signal Processing Reference
In-Depth Information
Figure 20.12
Illustration of a signal waveform distorted by faults
what is needed for reconstruction of data damaged as a result of system faults.
Therefore application of this technology for the development of sensor signal
processing systems tolerant to faulty signal sample values or to sensor failures is
well worth considering.
To demonstrate the potential of the algorithms discussed above for reconstruc-
tion of signals heavily distorted by fault bursts, a typical case of signal distortion
by faults was simulated and the obtained results follow. It was assumed that as
a digital sensor signal is corrupted by a powerful interference and many of the
signal sample values are completely distorted they had to be taken out of this sig-
nal at time instants indicated in Figure 20.12(a) by zeroes. The signal waveform
distorted by these fault bursts is shown in Figure 20.12(b).
The iterative signal reconstruction algorithm adapted to the specific fault se-
quence was used for recovery of the original signal. A zoomed segment of the
reconstructed signal sample value sequence is given in Figure 20.13. It can be
seen that the quality of the distorted signal reconstruction in this case is good as
the sample values of the reconstructed signal overlap the original signal. As ex-
pected, the algorithms for reconstruction of signals from their nonuniform sample
value sequences adapted to the sampling nonuniformities could be successfully
used also for recovery of signals distorted by system faults.
