Underwater acoustic communications play a critical role in ocean exploration, environmental monitoring, and underwater robotics. However, reliable communication remains challenging due to multipath propagation, Doppler effects, ambient noise, and the time-varying nature of underwater acoustic channels.
This project aims to investigate machine learning (ML) approaches to improve the reliability of underwater acoustic communication systems. The student will explore and implement ML techniques for tasks such as channel estimation, signal detection, and equalization. Using Python-based simulations and experimental data, the student will evaluate the performance of ML-based methods and compare them with conventional digital signal processing techniques.
The project offers an opportunity to gain hands-on experience in digital signal processing, wireless communications, and machine learning, while contributing to research on next-generation underwater communication technologies.
School
Electrical Engineering and Telecommunications
Research Area
Machine learning | Deep learning | Digital communications
Suitable for recognition of Work Integrated Learning (industrial training)?
No
- Research environment
- Expected outcomes
- Supervisory team
- Reference material/links
The research will be conducted in the Wireless Communications Laboratory at UNSW, a world-class facility at the forefront of wireless communications, signal processing, and information theory. The lab is equipped with high-performance workstations, GPUs, and software-defined radio platforms, enabling students to explore both theoretical concepts and practical system implementation. Students will have the opportunity to experiment with modern machine learning tools, apply AI-driven approaches to communication systems, and see how abstract theories are realized in practice. The lab supports flexible experimentation through both simulations and real-world testbeds, bridging the gap between theory and application.
The laboratory is led by Prof. Jinhong Yuan, Head of School, and includes a collaborative team of four postdoctoral researchers and engineers, alongside ten PhD, master’s, and undergraduate research students. Research outcomes from the lab are regularly published in leading journals and conferences, providing students with insight into high-impact research practices. This environment fosters close mentorship, teamwork, and active participation in ongoing projects, giving students hands-on experience and a strong foundation in both communications and AI-based methods.
An end-to-end emulation package demonstrating improved data rate and symbol error rate performance over the existing signal processing based transceiver architecture.
- Pattern Recognition and Machine Learning by Christopher M. Bishop
- Digital Communications by John Proakis