Abstract: Objectives Due to the unstable performance of traditional multi-channel automatic identification system (AIS) blind signal separation algorithms under low signal-to-noise ratio (SNR) conditions, and the fact that in actual communications, the number of source signals is often unknown, this paper proposes an improved joint approximate diagonalization of eigenmatrices (JADE) algorithm based on preprocessing to enhance robustness and stability. Methods First, the received signal is preprocessed using the singular spectrum analysis (SSA) algorithm to reduce noise. Next, the minimum description length (MDL) algorithm is used to estimate the number of source signals in the processed mixed matrix. Finally, to address the shortcomings of the traditional JADE algorithm, an improved JADE optimization algorithm is proposed to separate AIS mixed signals. Results By simulating the improved JADE algorithm, its performance was compared with that of the traditional JADE algorithm, the Fast Independent Component Analysis (FastICA) algorithm, the Robust Independent Component Analysis (RobustICA) algorithm, and the Information Maximization (Informax) algorithm. The results show that when separating 2, 3, and 4-channel AIS observation signals, the improved JADE algorithm achieves the highest correlation coefficient, indicating the strongest correlation between the separated signal and the source signal, and this value exceeds 0.8015 across all SNR ranges from 0 to 20dB. Additionally, in terms of crosstalk interference suppression, this method achieves the best suppression performance with values below 0.164 across all SNR ranges from -10 to 20dB. Conclusions The improved JADE optimization algorithm has significant advantages in terms of separation accuracy, algorithm reliability, and computational stability. The research results provide a theoretical basis and technical direction for improving the separation performance and real-time capabilities of AIS receivers and optimizing the design of maritime communication systems in actual engineering applications.