Cross-correlation tracking for Maximum Length Sequence based acoustic localisation

While Maximum Length Sequence based cross-correlation and localisation methods are specifically robust with respect to multiple forms of disturbances, the method is still subject of becoming 'side-tracked' by echoes, reverberations and other resonance effects. This article details a method to track the correct reading throughout a series of measurements which is specifically designed for the cross-correlation localisation technique as deployed in underwater environments by the Serafina project. The method is based on the selection of the most likely candidate rather than Kalman filters or other low-pass filters and extrapolation methods. Therefore every individual reading is an actual measurement, rather than an extrapolation. As transducers with non-linear frequency responses are common in energy-efficient underwater acoustic setups, the resulting signal deformations are explicitly addressed and corrected by means of inverse frequency transformations. The article discusses in detail a series of experiments performed in a highly reverberant underwater environment. The achieved performances include standard deviations of a less then 3 degrees in azimuth estimation, and 11 cm in range estimations throughout the presented sequence of experiments.