Background noise estimation using an expectation maximization algorithm

摘要

Measuring the background noise in the ocean can be challenging. Prior approaches to using measured data to estimate the background noise field have relied on collecting data during a series of tow vessel maneuvers. This paper presents an alternate approach to estimating the noise field that does not rely on tow vessel maneuvers to collect the necessary data. This approach fuses data from a traditional towed array with data from two spatially oversampled, small aperture planar arrays using an Estimation Maximization algorithm. The estimated noise field can be used as an input to various detection algorithms to improve performance.