Robust matched field processing via second order cone optimization

Abstract

Matched field processing on passive localization is prior to the classical passive source localization algorithm, because it is a model-based technique that use the environmental information, and has been extensively used for locating acoustic source in a multi path shallow water wave guide. But the principal issues are low resolution of localization, high side lobe level, and sensitivity to environmental mismatch. So an approach called robust matched field processing via second order cone optimization is proposed. We reformulate the matched field processing localization problem to be an optimization problem with accounting for white noise and side lobe constrains, and build a robust localization model. To evaluate the algorithm, some simulation experiments are processed in a pekeris shallow water environment. The results show that the proposed method can localize the acoustic source successfully even when some environmental mismatch exists.