Doctoral Defence: Yuan LIU

Members of the defense committee:

Prof. Dr. Fabrizio PASTORE, University of Luxembourg, Chairman

Prof. Dr. Bhavani SHANKAR MYSORE RAMARAO, University of Luxembourg, Supervisor

Dr. Mohammad ALAEE-KERAHROODI, University of Luxembourg, Member

Dr. Markus RUPP, Vienna University of Technology, Austria, Member

Dr. Thomas Stifter, IEE S.A., Luxemburg, Member

Abstract:

With the proliferation of sensing applications, including intelligent Internet of Things (IoT) and Integrated Sensing and Communication (ISAC) networks, the need for high accuracy sensing has become increasingly critical. This demand shifts radar applications from traditional long-range scenarios toward complex short-range environments, where channel modelling becomes critical for sensing tasks. Channels are determined by electromagnetic wave propagation within the environment. However, unlike communication channels, radar channel modelling needs to consider dynamic targets and their interactions with the surroundings. Sensing, in turn, is an inverse operation of modelling, i.e., it relies on received channel data and parametric models to estimate target locations and even reconstruct the environment. Moreover, radar is inherently an active sensing system, it can leverage specifically designed waveforms and Radio Frequency

(RF) configurations to effectively probe and adapt robust parameter estimation algorithms to channel characteristics. This thesis addresses these requirements through three interrelated parts, with the channel model serving as a bridge between the environment and the sensing system. Given a known digital map and target dynamics, Part I develops a sensing channel modelling and simulation platform to generate short-range Multiple Input Multiple Output (MIMO) radar

signals. Reciprocally, Part II and III focus on leveraging the received channel data and parametric models for target localization and environment mapping. Specifically, Part II investigates the emerging challenges, e.g., multi-bounce propagation, Near Field (NF), and Spatial Non-Stationarity (SNS) effects, and exploits these phenomena for localization and environment reconstruction. Part III addresses the parameter estimation ambiguity in time-varying channels, and proposes a novel Code Division Mode Multiplexing (CDMM) vortex waveform and Doppler-robust algorithm, which enhances sensing performance under dynamic conditions and benefits ISAC applications.