Improvement of Acoustic Signal Detection Using Dual-Tree Complex Wavelet Transform

An essential component of many sonar applications is signal detection in the water. However, signal detection can be challenging due to colored noise that deviates from the Gaussian distribution. This can lead to well-known devices like Matched Filters (MF) functioning badly, which reduces their utility for identifying signals from below the surface. In order to improve underwater detection, the Dual-Tree Discrete Wavelet Transform (DT-DWT), also known as a complex wavelet transform, is recommended. DT-DWT enables level-dependent thresholding, which can enhance the MF's signal detection capabilities by removing the underwater noise. The proposed methodology has demonstrated superior performance compared to a commonly used wavelet transform de-noising method. In order to investigate whether the recommended method is superior to other detection techniques like the Discrete Wavelet Transform (DWT), the Local Optimum detector (LO), and the traditional Match Filter (MF), the study also looks at how well the suggested method performs in comparison to these techniques. At an Energy-to-Noise Ratio (ENR) of 7.32 dB for MF, 4.15 dB for LO, 4.05 dB for DWT, and 3 dB for the proposed DT-DWT, the detectors have achieved a 90% probability of detection with a false alarm probability of 0.1.