The problem of detecting anomalous noise in the case of a set of continuous and discrete observations with memory is considered. The potential properties of the algorithm relative to the probabilities of errors of omission or of false detection are investigated on the basis of the Kullback divergence. In one special case the efficiency with which anomalous noise is detected in a channel with memory relative to a channel without memory is investigated.
|Number of pages||12|
|Journal||Automatic Control and Computer Sciences|
|Publication status||Published - 26 Aug 2004|
ASJC Scopus subject areas
- Control and Systems Engineering
- Signal Processing