A method for compact data storage in optimization of k-stage detectors

A method is developed for the compact storage of statistically simulated da ta that extends the method for optimizing parameters of k-stage signal dete ctors to the case when the probability of false decisions is low. Algorithm s for dividing a set of statistic trajectories into groups, marking borders between trajectory sets, and extracting a trajectory from a group, which l ay behind the proposed method, are described. The method is used to optimiz e and analyze the k-stage detectors for coherent (with a random initial pha se) and incoherent pulse bursts. Results are provided that compare the k-st age detectors with detectors operating by fixed-length samples.