Adaptive CFAR Tests for Detection of a Signal in Noise and Deflection Criterion

Abstract

Signal detection is the area of study that deals with the processing of informationbearing signals for the purpose of extracting information from them. Specific applications in which detection techniques are useful include radar, seismology, radio astronomy, sonar, speech and image processing, automatic control, medical signal processing, and optical communications. In general, detection applications involve making inferences from observations that are distorted or corrupted in some unknown manner. Moreover, the information that one wishes to extract from such observations is a fortiori unknown to the observer. Thus it is very useful to cast detection problems in a probabilistic framework in which unknown behavior is assumed to be random. In this light, detection theory fits properly within the province of statistical inference, and this is the interpretation to be used throughout this treatment.