Quasioptimum Algorithm for Detecting Narrow-Band Signals against a Random Noise Background

Abstract

A coherent averaging technique is suggested for detecting and estimating parameters of narrow-band signals against an additive random noise background. Expressions have been obtained for the mean value and variance of the spectral density estimate. The algorithm is compared with those of the optimum filtering and incoherent averaging methods. Results are presented on numerical simulation and application of the algorithm to detection of the coherent ELF signals produced by demodulation of a powerful high frequency AM radiation due to nonlinear properties of the ionospheric plasma.