/Enhancing Radio Frequency Radar Data Interpretability through Explainable AI Techniques

Enhancing Radio Frequency Radar Data Interpretability through Explainable AI Techniques

Gent | More than two weeks ago

Understand and explain how machine learning techniques interpret radio frequency data for radar applications

The rapid advancements in radar technology have led to radar innovations in various applications, including advanced driver assistance, autonomous vehicles, activity recognition, vital sign monitoring and environmental monitoring. Novel machine learning techniques are often used to train radar solutions based on large radar datasets containing annotated time series from pulse radars, mmwave radars of UWB radars. However, the black-box nature of traditional machine learning algorithms often hinders the interpretability of radar-based decision-making systems. This PhD research aims to bridge the gap between radar applications and human understanding by investigating Explainable Artificial Intelligence (XAI) techniques tailored for radar data.


Expected Outcomes:

This research will contribute to the field of Radio Frequency (RF) radar applications and artificial intelligence by providing tools and methodologies to enhance the interpretability of radar-based systems.

This includes

  • Developing Radar-Specific XAI Models: Design novel XAI models and techniques customized for radar data, considering the unique characteristics, such as high-dimensional data, temporal dependencies, and noise patterns.
  • Interpretable Feature Extraction: Investigate methods for extracting interpretable features from radar data that capture relevant information while reducing complexity.
  • Real-time Explainability: Explore real-time or near-real-time XAI solutions for radar applications to support decision-making in dynamic environments.
  • Transferability to various radar domains: Apply the developed XAI techniques to predict and enhance transferability to diverse radar applications, such as target tracking, object recognition, and vital sign monitoring, to assess their generalizability and adaptability.


The PhD dissertation will offer a comprehensive framework for integrating explainable AI techniques into radar applications, improving decision-making, reducing false alarms, and enhancing human-machine collaboration in radar-intensive environments. By shedding light on the decision-making process of radar systems, this research will pave the way for safer and more reliable applications of radar technology in critical domains.


The successful PhD candidate will be part of a large IMEC team working on the design and implementation of AI/ML for wireless communications and networks. This is a unique opportunity to develop innovative, multi-disciplinary technology and share future wireless networks. You will publish your research in top-tier journals and conferences.

Required background: Engineering Technology, Engineering Science, Computer Science, Informatics or equivalent

Type of work: 70% design/modeling/simulation, 20% experimental and 10% literature

Supervisor: Eli De Poorter

Co-supervisor: Steven Latre

Daily advisor: Adnan Shahid

The reference code for this position is 2024-151. Mention this reference code on your application form.

Who we are
Accept marketing-cookies to view this content.
Cookie settings
imec's cleanroom
Accept marketing-cookies to view this content.
Cookie settings

Send this job to your email