Session 3.4c Tutorial: Trainable Radios: Exploring the Potential of Differentiable Architectures for Communication Systems and Electronic Warfare
Tracks
| Thursday, November 16, 2023 |
| 11:30 AM - 12:30 PM |
| Menzies Theatre |
ㅤ
Advancements in artificial intelligence (AI) and machine learning have led to significant breakthroughs in various fields, such as computer vision and natural language processing. A key factor behind the success of these AI applications is the use of differentiable neural architectures. This presentation explores the application of these architectures in developing "trainable radios" for communication systems and electronic warfare (EW).
Within the Defence sector, communication systems and EW play pivotal roles. These domains are renowned for their complexity and ever-changing nature, presenting distinct challenges. Differentiable architectures are uncommon in communications and EW, most likely due to the established success and robust benchmarks of conventional engineering techniques. The talk seeks to uncover the untapped potential of trainable radios, offering versatile solutions for low-probability-of-intercept communications, channel sharing, jamming, and more.
Speaker/s
Mr Daniel Gibbons
Machine Learning Engineer
DEWC Services
Daniel Gibbons is a Machine Learning Engineer at DEWC Services, who holds an MPhil in Electrical and Electronic Engineering from the University of Adelaide with a specialisation in Machine Learning. Daniel's background spans various domains, including deep learning, control theory, and multi-agent systems. In recent years, Daniel has been developing deep learning approaches tailored for SIGINT and communications applications. Daniel is currently working with DST Group, providing valuable machine learning expertise for developing enterprise ModelOps. This initiative aims to accelerate the integration and improvement of AI models in Defence applications.
