A Robust Sub-Pulse Multipath Clusters Accumulation Detection Method for Underwater Acoustic Signals

Because of the complex characteristics of underwater acoustic channels, including severe multipath propagation, time-varying Doppler shifts, and high noise levels, underwater acoustic signal detection faces significant challenges. Under these conditions, traditional matched filtering suffers from performance degradation, particularly in the presence of nonlinear Doppler effects caused by variable-speed motion. In this paper, a robust sub-pulse cluster accumulation (SPCA) detection method is proposed, which decomposes the signal into sub-pulses for independent Doppler processing. Using the sparse clustering characteristics of underwater acoustic channels, the proposed method applies a dual-threshold energy screening approach to detect clusters and accumulates multipath energy to suppress noise. The fusion detection criterion requires consistent detection results across multiple sub-pulses, further enhancing robustness. Experimental results demonstrate that at a false alarm rate of 1% and a signal-to-noise ratio (SNR) of −16 dB, the SPCA detection probability exceeds that of sub-pulse matched filtering by 16%, significantly outperforming conventional methods. These findings validate the effectiveness of exploiting channel sparsity and sub-pulse fusion for reliable detection in underwater environments.