Sparsity-Driven ALE Algorithm of Underwater Acoustic Tonals with Experiment Verification

The tonal components in ship radiated noise often contain rich target feature information, so it can be used as the main signal for passive sonar detection. However, these tonal components are typically submerged in the ambient noise. Adaptive line enhancers can suppress background noise and enhance the signal-to-noise ratio of the tonal components. In this letter, the frequency-domain driven Zero-Attracting Adaptive Liner Enhancer (ZA-ALE) algorithm is exploited to analyze sea trial data in the South China Sea. Compared with conventional ALE, ZA-ALE can obtain at least 6 dB extra SNR gain with simulation data and 1.49 dB extra SNR gain with sea trials data. The results demonstrate that the ZA-ALE algorithm exhibits good noise suppression capabilities.