Toward Intelligent Underwater Acoustic Systems: Systematic Insights into Channel Estimation and Modulation Methods

Underwater acoustic communication (UWA) is essential for many mission-critical applications such as such as deep-sea exploration, maritime security, and environmental monitoring. However, it continues to face several challenges from multipath propagation, channel variations, and unpredictable underwater noise. Recent advancements in artificial intelligence, particularly machine learning (ML) and deep learning (DL), have significantly improved channel estimation, adaptive modulation, and modulation recognition. To assess these advancements, a systematic literature review (SLR) is needed. Existing review articles focus on isolated parts like channel estimation or modulation recognition. It implies that they don’t compare methods across domains, making it hard to understand their overall impact in real applications. Moreover, they often miss system-level comparisons and practical constraints such as bandwidth, system complexity, and real-time adaptability, making it hard to judge real-world significance. To bridge this gap, this SLR provides a comprehensive and structured synthesis of 43 recent studies (2020–2025), covering both single-carrier and multi-carrier UWA systems. It categorizes ML/DL techniques applied at the physical layer and standardizes performance metrics such as bit error rate, training loss, and computational overhead. Furthermore, it highlights key architectural trends in model design and synthesizes insights across diverse scenarios to identify existing research gaps. This level of integration and comparative analysis has not been presented in previous reviews. As a result, the holistic perspective offered by this SLR serves as a timely and valuable resource for guiding future advancements in robust, intelligent, and scalable UWA communication systems.