Extreme accessible probabilities of true recognition of multi-dimensional vector signals
Y.A. Furman and A.A .Rozhentsov
Yoshkar-Ola
Pages: 29–41
Abstract
The problem of definition of extreme accessible probabilities of true signal recognition by means of an ordered set of k-dimensional vectors is solved. The choice of a space for signal presentation, the criterion of decision making, the kind of reference signals of the alphabet and its volume, as well as the model of a signal recognized are substantiated. An expression for calculating the extreme accessible probability of true recognition of a k-dimensional signal is obtained as a -multiple definite integral of the normal probability distri- bution density. Validity of analytically obtained results is experimentally verified.
|
