DETECTION OF WEAK STOCHASTIC SIGNALS IN DOUBLE-SIDED EXPONENTIAL NOISE USING GENERALIZED ENERGY DETECTOR

Many locally optimum, suboptimum, and ad hoc detectors exploit knowled ge of the noise probability density function (pdf) to obtain their tes t statistic for the detection of stochastic signals in non-Gaussian no ise. These detectors appear to offer potential for highly improved per formance for very small signal to noise ratios (asymptotically approac hing zero). This paper studies a locally optimum detector namely the g eneralized energy detector, by deriving analytical expressions for a p erformance measure using a double-sided exponential noise pdf. The pap er concludes that the apparently large gains predicted by the asymptot ic theory may not be reachable within the practical range of signal-to -noise ratios.