Abstract:As the Internet-of-Things (IoT) continues to expand and reshape our world, new vertical application scenarios have emerged, such as underwater communications, leading to increased interest in academia and industries. The multiple-input multiple-output (MIMO) technology plays a critical role in enhancing channel capacity for underwater acoustic (UWA) communications, where accurate channel prediction is essential for system performance. In this article, we propose a novel efficient channel impulse response (CIR) prediction model for the UWA MIMO communications with a small adaptive bidirectional gated recurrent unit (ABiGRU) network. The proposed model can capture the channel information without additional knowledge of the internal properties of the channel itself. Moreover, it first utilizes preceding short-term CIR data from the channel estimation for online training, and then exploits the trained model for the CIR prediction, which tracks time-varying UWA channels. To verify the effectiveness of the predicted CIRs, we design a scheme combining a space-time block coding (STBC) and MMSE pre-equalization for the UWA MIMO system. Our proposed STBC-MMSE pre-equalization scheme has demonstrated practical feasibility and low bit-error rate (BER) in numerical simulations. In addition, we evaluate the prediction error performance of the proposed ABiGRU network through comparison with the widely used minimum mean square error (MMSE) algorithm and two common recurrent neural networks (RNNs) predictors, i.e., the GRU and long short term memory (LSTM) network. Finally, we conduct realistic in-field UWA MIMO experiments to demonstrate and justify the superiority of the proposed ABiGRU network, which can lay the solid foundation for cost-effective UWA MIMO communications for building promising underwater IoT sensor networks.

A Multi-Task Learning Framework for Underwater Acoustic Channel Prediction: Performance Analysis on Real-World Data

Channel Prediction for Underwater Acoustic Communication: A Review and Performance Evaluation of Algorithms

Multi-Task Learning Based Underwater Acoustic OFDM Communications

Effective Underwater Acoustic Target Passive Localization of Using a Multi-Task Learning Model with Attention Mechanism: Analysis and Comparison under Real Sea Trial Datasets

Advancing Robust Underwater Acoustic Target Recognition Through Multitask Learning and Multi-Gate Mixture of Experts.

Adversarial Multi-Task Underwater Acoustic Target Recognition: Toward Robustness Against Various Influential Factors.

Advancing Robust Underwater Acoustic Target Recognition through Multi-task Learning and Multi-Gate Mixture-of-Experts

Time-Varying Channel Prediction for RIS-Assisted MU-MISO Networks Via Deep Learning

Underwater Acoustic Adaptive Modulation with Reinforcement Learning and Channel Prediction

Experimental Results of Underwater Sound Speed Profile Inversion by Few-shot Multi-task Learning

Machine Learning for Underwater Acoustic Communications

Adversarial multi-task underwater acoustic target recognition: towards robustness against various influential factors

Online machine learning-based channel estimation for underwater acoustic communications

A Novel Multi-Feature Fusion Model Based on Pre-Trained Wav2vec 2.0 for Underwater Acoustic Target Recognition

Underwater target recognition based on adaptive multi-feature fusion network

Adaptive Modulation and Coding in Underwater Acoustic Communications: a Machine Learning Perspective

Channel Prediction Using Adaptive Bidirectional GRU for Underwater MIMO Communications

Multi-Features Fusion for Underwater Acoustic Target Recognition Based on Convolution Recurrent Neural Networks

Channel State Information Prediction for Underwater Acoustic Multiple-Input Multiple-Output Communications

A Multi-Task Network: Improving Unmanned Underwater Vehicle Self-Noise Separation Via Sound Event Recognition

Data-Driven Signal Detection for Underwater Acoustic Filter Bank Multicarrier Communications