Journal Article
Long Short-Term Memory Neural Network Assisted Peak to Average Power Ratio Reduction for Underwater Acoustic Orthogonal Frequency Division Multiplexing Communication
A KSII Transactions on Internet and Information Systems article on LSTM neural networks for reducing peak-to-average power ratio in underwater acoustic OFDM communication.
Abstract
The underwater acoustic wireless communication networks are generally formed by the different autonomous underwater acoustic vehicles, and transceivers interconnected to the bottom of the ocean with battery deployed modems. Orthogonal frequency division multiplexing (OFDM) has become the most popular modulation technique in underwater acoustic communication due to its high data transmission and robustness over other symmetrical modulation techniques. To maintain the operability of underwater acoustic communication networks, the power consumption of battery-operated transceivers becomes a vital necessity to be minimized. The OFDM technology has a major lack of peak to average power ratio (PAPR) which results in the consumption of more power, creating non-linear distortion and increasing the bit error rate (BER). To overcome this situation, we have contributed our symmetry research into three dimensions. Firstly, we propose a machine learning-based underwater acoustic communication system through long short-term memory neural network (LSTM-NN). Secondly, the proposed LSTM-NN reduces the PAPR and makes the system reliable and efficient, which turns into a better performance of BER. Finally, the simulation and water tank experimental data results are executed which proves that the LSTM-NN is the best solution for mitigating the PAPR with non-linear distortion and complexity in the overall communication system.
Plain-Language Summary
This work uses machine learning to improve underwater acoustic communication, helping OFDM systems transmit data more efficiently in difficult underwater environments.
Why This Paper Matters
Underwater acoustic communication systems often rely on battery-powered modems and operate under severe bandwidth, multipath, Doppler, and energy constraints. High PAPR in OFDM systems increases power consumption, nonlinear distortion, and BER. This paper shows how LSTM-based machine learning can reduce PAPR and support more energy-efficient underwater acoustic modems and transceivers.
Research Summary
This paper focuses on underwater acoustic communication, where reliable data transfer is difficult because the channel is noisy, bandwidth-limited, and affected by environmental conditions. OFDM is attractive for underwater communication, but high peak-to-average power ratio can reduce efficiency and create practical transmission challenges.
The study applies long short-term memory neural networks to help reduce PAPR in underwater acoustic OFDM systems. The use of LSTM models is motivated by their ability to learn temporal patterns, which can be useful in communication signals affected by sequence-dependent behavior.
The broader contribution is the application of machine learning to a signal-processing problem in underwater networks. This supports more efficient communication for systems such as autonomous underwater vehicles and battery-powered acoustic modems.
LSTM-Assisted Underwater Acoustic OFDM Framework
OFDM Signal Generation
Generates underwater acoustic OFDM symbols for transmission through a multipath and Doppler-affected UAC channel.
PAPR Estimation
Measures high-energy peaks in OFDM symbols using PAPR and CCDF analysis before transmission.
LSTM-NN Reduction
Uses an LSTM neural network with memory to select lower-PAPR signal patterns and reduce nonlinear distortion.
BER and Energy Evaluation
Evaluates received constellations, BER, and energy efficiency using simulation and water-tank experimental data.
Key Contributions
- Applies LSTM neural networks to PAPR reduction in underwater acoustic OFDM communication.
- Targets communication systems used by autonomous underwater vehicles and battery-powered modems.
- Connects machine learning with practical signal-processing challenges in underwater networks.
Modeling Approaches Reviewed
LSTM Neural Networks
Uses recurrent memory to identify and reduce high PAPR patterns in OFDM signal sequences.
Connectionist Temporal Classification
Supports sequence classification for high-peak and low-peak patterns without requiring explicit alignment.
BELLHOP Channel Modeling
Simulates underwater acoustic propagation effects for evaluating signal transmission and receiver performance.
16-QAM OFDM Evaluation
Uses 16-QAM modulation to study transmitted and received OFDM symbols under underwater acoustic channel conditions.
Energy Efficiency Analysis
Compares power and energy efficiency behavior across LSTM-NN, DNN autoencoder, NN-ACE, and traditional OFDM schemes.
Research Gaps
Publication Details
- Type
- Journal Article
- Venue
- KSII Transactions on Internet & Information Systems
- Year
- 2023
- Volume
- 17
- Issue
- 1
- Pages
- 239-260
Authors
Research Topics
Links and Access
Citation
@article{raza2023long,
author={Raza, Waleed and Ma, Xuefei and Song, Houbing and Ali, Amir and Zubairi, Habib and Acharya, Kamal},
title={Long Short-Term Memory Neural Network Assisted Peak to Average Power Ratio Reduction for Underwater Acoustic Orthogonal Frequency Division Multiplexing Communication},
journal={KSII Transactions on Internet and Information Systems},
year={2023},
volume={17},
number={1},
pages={239--260},
doi={10.3837/tiis.2023.01.013}
}