R. J. Polge

Principal Investigator

R. J. Polge

Contract Number

Grant No. DA-AMC-01-021-64-G1, Task A

Contract Name

Grant No. DA-AMC-01-021-64-G1, Task A


U. S. Army Missile Command

Date of Report


Date of Project


School/Organization Name

University of Alabama Research Institute

Subject Area

Radar detection

Document Type



This report studies the two-steps detection of search radar pulses in the presence of noise. The two informations, location and existence of a target, are obtained simultaneously in a conventional radar receiver, but successively in a two-steps radar receiver. The location of all the significant pulses, i.e. targets and large noise, is obtained first by a low level continuous threshold detection. The first threshold detector unlatches a second threshold detector (once for each significant pulse) which decides whether the significant pulse is a target pulse or a noise pulse. Once the location of the pulses in known, the problem of radar becomes a problem of telemetry and the adaptive decision technique developed in the UARI Report No. 33 [ 1 ] for the detection of pulse code modulated signals is applicable. The noise just before the unknown signal, determined by sampling, and the noise during the interval of detection are more or less correlated; therefore, using correlation techniques, it is possible to predict the noise anywhere during the interval of detection. The probability of error in the detection is considerably reduced in an adaptive scheme where the noisy signal is corrected by substracting the predicted noise. The mathematics are more difficult for the adaptive two-threshold detection than for the adaptive PCM detection because of the complexity of the joint probability density after envelope detection. The probability of false alarm for an adaptive two-steps detection, P_fat' is expressed by a double integral. P_fat is expanded as a power series in powers of p^2 where p is the autocorrelation coefficient between sampled and detected signal; the series is valid only for small values of p. Finally, P_fat is computed with a digital computer for typical value of p and D_C (level of the second threshold detector). It is shown that the probability of false alarm is smaller for an adaptive two-steps detection than for a conventional detection, especially for large values of p.

Publication Date



Final Technical Report

Citizenship Confirmation


Sensitive Material Confirmation




To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.