A TWO-STAGE LIKELIHOOD-RATIO DETECTION PROCEDURE FOR AN UNKNOWN WAVEFORM WHICH IS RANDOMLY REPEATING IN NOISE.
CARNEGIE INST OF TECH PITTSBURGH PA GRADUATE SCHOOL OF INDUSTRIAL ADMINISTRATION
Pagination or Media Count:
Suppose we have a stochastic process which is made up of an unknown waveform of finite duration, which is repeated randomly in Gaussian noise with a known covariance function. The rate of recurrence of the waveform is a known constant, and the signal-to-noise ratio of the waveform in the noise is small. A two-stage procedure is developed for detecting the epochs of the waveform based upon efficient large-sample estimates of the autocorrelation of the waveform. The error probabilities are given for this procedure which is based upon approximation of the likelihood-ratio. A bound for the error due to the approximation of the likelihood function is given. Author