Multipath compression by multiple convolutions

Blind deconvolution algorithms can be useful as pre‐processors for signal classification algorithms in shallow water. These algorithms remove the distortion of the signal caused by multipath propagation when no knowledge of the environment is available. A very simple blind algorithm is presented here which utilizes convolutions of multiple data channels for convolutional smoothing of multipath. Computer simulation studies show significant performance improvement when propagation effects are mitigated by the use of this blind technique. [Work supported by ONR and the Naval Research Laboratory.]