Acoustic data preprocessing in the TOMOLAB/OCTOPUS project

Abstract

This paper presents the acoustic data preprocessing task in the TOMOLAB graphic interface tool. TOMOLAB is a goal of the OCTOPUS project, under the MAST III program of the European Community. Data preprocessing consists in transforming raw tomographic data into compressed/correlated peaklike data. It is the first step before time delay estimation, time delay identification, and time delay inversion. The complete data processing must perform matched filtering, multisequence averaging, Doppler/phase compensation, and mooring motion correction. A synthetic view is given of several technical choices that integrate these tasks, jointly or successively. This work is based on internal studies and existing pieces of literature. The Hadamard transform, the Fourier matched filtering, fast Doppler processing, phase estimation/correction algorithm, delay shifting in the time or Fourier domain, data format and sampling are discussed. Efficient data processing has to account for the computing speed, the compression ratio, and achievable accuracy of delay estimation.