Subject

Integrated sensing and communication (ISAC) uses a wireless communication channel, for example WiFi or cellular networks, to also sense the environment through more elaborate processing of channel measurements. This dual use of the same hardware increases spectral and energy efficiency, and allows us to combine sensing and communication functions into a single system, offering applications in various fields. ISAC improves efficiency, accuracy, and real-time data processing, benefiting traffic management, remote monitoring, precision farming, and more by integrating sensing data with communication capabilities.

ISAC requires signal processing methods for signal reconstruction and AI methods for data analysis. This thesis will develop interpretable by design deep learning methods for signal processing and AI for ISAC.

Deep unfolding is a method to design interpretable AI models by unrolling an optimization algorithm and mapping the (sub)steps to corresponding neural network layers to obtain a machine learning model that incorporates the domain knowledge from the original algorithm into its architecture. This approach results in very compact and efficient models, with many use cases in signal and image processing.

Kind of work

The student will use compressed sensing techniques to extract useful information from the (sparse) channel measurements. The goal of this project is for you to explore the state-of-the-art in ISAC and get acquainted with the typical data types and processing steps. You will select or put together a baseline ISAC dataset from what is available in the literature. Next, the task is to adopt the deep unfolding models from our research group to compressed sensing for ISAC, and apply them to tasks like environment sensing, localization or activity recognition. The goal is to show the benefits of deep unfolding models for ISAC processing, and compare them to the state-of-the-art.

Framework of the Thesis

The thesis will obtain access to unique data for human activity monitoring and environment sensing in smart cities and will design AI and signal processing models based on deep unfolding. Support will be provided in the development of models given the vast experience of the group in designing interpretable by design models for signal processing and AI tasks. The thesis will also explore collaborations with imec.

Number of Students

1 or 2

Expected Student Profile

Good knowledge of machine learning and signal processing. Good knowledge of Python (including PyTorch). Understanding communication signals and data is also a plus.