Cross-range imaging of a cooperative source using passive inverse synthetic aperture technique

Abstract

In recent literature on underwater acoustics, great emphasis has been placed on the impact of anthropogenic sound on marine life. The monitoring and reduction of noise emissions from submarines and ships has become an integral aspect in manufacture and maintenance processes. The sound produced by large ships is known to contain tonals, which are generated by various machines operating on the marine vessel. The objective of this paper is to perform a spatial mapping of the tonal sources in cross-range by employing the concept of passive inverse synthetic aperture. A single stationary sensor observes the sound emitted by the cooperative source under uniform linear motion at close range over finite observation duration. It is assumed that the velocity and the range at the closest point of approach of the moving vessel, modeled as a distributed source, are known accurately. The emitted tonals undergo a time-varying Doppler shift due to the relative motion of the source as compared to the stationary sensor. We propose a two-step algorithm for cross-range imaging of the cooperative source, which involves the estimation of the tonal rest frequencies followed by the estimation of the time at the closest point of approach corresponding to each of the estimated tonal frequencies. The prominent tonal frequencies are estimated using a novel one-dimensional optimization based on time warping of the received signal. Then, matched filters are designed to determine the time at the closest point of approach corresponding to each of the Doppler shifted tonal frequencies, which indicate the location of the tonal sources relative to each other in the cross-range. The proposed algorithm is also capable of distinguishing two or more sources that are spatially displaced, but emit the same tonal frequency. The issues related to the design of the matched filters in the proposed algorithm and the resolution achievable in the cross-range are discussed. Simulations are used to assess the performance of the proposed algorithm in presence of noise. The proposed algorithm is also applied on an actual underwater recording.