Application Cases

Experiment Name: Research on the Principles and Methods of Broadband Underwater Acoustic Array Signal Processing

Research Direction: Underwater Communication

Test Equipment:
Signal generator, noise generator, ATA-2021B high-voltage amplifier, preamplifier, programmable filter amplifier.

Experimental Procedure:
The experiment was conducted in an anechoic water tank with dimensions of 20.0 m (length) × 8.0 m (width) × 7.0 m (depth). The noise source was positioned approximately 2.0 m away from the tank wall (width direction). A baffle, measuring about 1.5 m × 1.5 m, was placed directly in front of the noise source (to the right in the diagram), about 25 m from the tank wall. The receiving array was positioned approximately 0.5 m in front of the baffle. The signal source was located about 4.0 m directly in front of the receiving array, with the signal propagation direction forming an angle of about 5° relative to the normal of the receiving array. The noise source, baffle center, signal source, and receiving array were all submerged at a depth of approximately 2.5 m below the water surface. The signal source generated a sinusoidal signal, while the noise generated by the noise source propagated around the baffle and was received by the receiving array to simulate spatially white noise (note: the experimental noise model differs from the theoretical noise model).

Figure: Schematic Diagram of the Experimental Setup

The experimental configuration is shown in the figure above. The signal and noise were generated by two separate sources. One source produced a single-frequency sinusoidal signal at 3.0 kHz, while the other generated broadband Gaussian white noise, which was then filtered through a bandpass filter with a frequency range of 25 kHz to 45 kHz. After generation, both the signal and noise were amplified by a power amplifier and transmitted via a transmitting transducer.

Experimental Results:
Data were collected under four different signal-to-noise ratio (SNR) conditions: pure noise, SNR = 0 dB, SNR = 6 dB, and SNR = 12 dB. For each condition, 80,000 data points were recorded at a sampling frequency of 20.0 kHz.

Figures 6.4.3 and 6.4.4 show the waveform and spectrum of the received signal from one element of the receiving array under SNR = 6 dB.

The collected data were divided into five non-overlapping segments of 1,024 points each. Direction of arrival estimation was performed for each segment using the signal phase matching principle. The estimation results under different SNR conditions are summarized in Table 6.4.1.

Theoretical analysis and simulation results indicate:

1. Under low SNR conditions, due to differences between the experimental noise model and the theoretical model, the variance of direction estimation using the multi-element array signal phase matching method is relatively high. However, the variance decreases rapidly as the noise level decreases.

2. Increasing the number of spatial array elements improves estimation accuracy under the same SNR conditions.

3. In terms of spatial azimuth spectra, the spectral peaks obtained using the signal phase matching principle with multi-element arrays are significantly sharper than those obtained with conventional beamformers.

Recommended Voltage Amplifier: ATA-2021B

Figure: Specifications of the ATA-2021B High-Voltage Amplifier