Non-Cooperative Multi-Target DOA Estimation for DS-CDMA Signals Using a Single Acoustic Vector Sensor

To address the requirements of marine information security and non-cooperative underwater acoustic communication networks, effective direction-of-arrival (DOA) estimation for non-cooperative Direct-Sequence Code Division Multiple Access(DS-CDMA) network nodes is critical. Conventional DOA estimation methods face challenges in wideband DS-CDMA scenarios due to the lack of prior information about signal sources. This paper proposes a matched correlation-based DOA estimation algorithm using an improved fast independent component analysis(Fast-ICA)method, which leverages the high directivity of a single Acoustic Vector Sensor(AVS) to achieve blind signal separation through optimized preprocessing, followed by matched correlation processing for azimuth estimation. Simulation results demonstrate that the proposed algorithm outperforms the weighted histogram and signal statistical methods in terms of accuracy and robustness, proving its effectiveness for underwater non-cooperative DS-CDMA target localization.