Extraction of signals from background interference with unknown parameters in quantized sample

Decision rules were constructed for detection of signals appearing with background interference in amplitude-quantized samples. The rules are to ensure that the false-alarm probability will not depend on the interference parameters and, therefore, the latter need not be a priori known. Samples with a finite number of realizations L = (M + 1) sub V are considered, including nonrandomized and randomized decision rules for acceptance of the signal present hypothesis through group and its subgroups identification on the basis of their sufficient statistics. The correct-detection probability is calculated for each case. Simple examples demonstrate that ignorance of the interference parameters is much less costly in the case of quantized samples than in the case of nonquantized ones.