Bottom-bounce ray angle subspace detector in the shadow zone of deep water

Passive source detection is a challenging problem in the shadow zone, where the sound is contributed primarily by bottom-bounce rays. The conventional beamforming detector (CBFD), which utilises the sound energy from a single direction, suffers potential significant performance degradation in the multipath-signal scenario. The matched field detector (MFD) offers optimal performance by exploiting full-wave field characteristics but is limited due to its reliance on prior ocean environmental knowledge. The authors demonstrate that the incident sound on a near-surface vertical line array in the shadow zone can be approximated as a coherent sum of two plane waves that share a symmetric arrival angle about the horizontal. This leads to the signal subspace depending only on the arrival angle, which the authors call the bottom-bounce ray angle subspace (BRAS). With generalised likelihood ratio test theory, the authors further derive the BRAS detector (BRASD). It can utilise the full signal energy under the premise of weak environmental knowledge requirements and is generally superior to the CBFD. Simulation results in a typical deep-water channel under different source-position and array configuration conditions demonstrate the effectiveness of the BRASD and suggest that it can even offer a performance comparable to that of the MFD under specific conditions.