Velocity filters for multiple signal extraction by multiple coherent interference attenuation in sensor array data

Abstract

The main objective of this work is to derive a vector of array filters for processing the recording of an array of sensors-geophones or hydrophones-based on a general model for the array data which incorporates multiple signal and multiple interference in addition to random noise. The signals and interferences have different arbitrary waveforms and no knowledge of these waveforms is assumed. The rationale followed is to give first priority-at each discrete frequency-for achieving all-pass conditions for the signals (if at all possible), and second priority for suppressing the interferences (if not in conflict with the first requirement) and a third priority for attenuating the random noise. Such an approach suits the typical case of array data dominantly contaminated by coherent interferences rather than random noise.