Narrowband signal detection in a reverberation-limited environment

Experimental work with three different adaptive processors has been carried out in order to study the problem of signal detection in a reverberation-limited environment. Single-channel real data as well as synthetic data have been used for the reverberation signals. Synthetic echoes and synthetic stepped FM pulses are regarded as the contact signals to be detected. Simulation results using an adaptive notch filter (ANF), an adaptive noise canceller (ANC) with its weights being adapted by means of the least-mean square (LMS) algorithm, and an ANC in the form of a least-squares lattice (LSL) filter, are presented. It is shown how the ANC, by means of the LSL, is capable of recovering even zero-Doppler echoes under certain signal conditions.