Digital data processing considerations in radar

The paper describes various aspects of radar processing techniques which are being studied in order to determine their effect on overall system performance. Many data extraction systems make decisions regarding the presence or absence of targets by comparing the amplitude of the envelope of the demodulator output with a threshold which depends usually on the prevailing noise level. Studies have been made to determine if this is an optimum procedure, by considering other functions of the demodulator output to be used in the detection process. Energy per unit time is a possible alternative as a detection parameter. Statistical methods are used to compare performance, involving the determination of probability distribution of the parameter used in the detection process. The distribution of some functions cannot be obtained by normal analytical methods; a computer method is described which determines an approximation to the required distribution from which an estimate of performance can be made. The use of sampling theory is indicated in the study of optimum sampling of the decision parameter and comparisons are made between equipment and performance of systems using the conventional sampling rate and systems using the value indicated by the sampling theory. The effect on performance of a digital integrator of quantizing the decision parameter into two or more levels is studied and compared with results using analogue waveforms.