A learning method for localizing objects in reverberant domains with limited measurements.

This article presents a learning (training)-based method for localizing objects in enclosures. Wave propagation in enclosures can lead to mixing of the wave energy, ultimately leading to incoherent spreading of information. This makes the localization problem challenging. However, spreading of the wave energy can lead to multiple interrogations of each point in the enclosure, which is in essence reminiscent of an ergodic or a closely ergodic behavior. Hence, any substructural changes in the enclosure can be sensed with sufficient information carried by the wave energy flow. Furthermore, temporal information buried in data makes it feasible to conduct only a few spatial measurements. A localization scheme is presented that benefits from the reverberant field and can reduce the required number of spatial measurements.