OFDM-Based Sensing: Signal Processing Strategies for Structural Vibration Monitoring and Enhanced Ranging

The structural integrity of civil infrastructure represents a critical challenge for public safety and maintenance management. This thesis investigates the use of OFDM radar sensing, within the Joint Communication and Sensing (JCAS) paradigm, to address two distinct sensing requirements: structural vibration monitoring for Structural Health Monitoring (SHM) and high-accuracy multi-target ranging. • For SHM applications, a bistatic slow-time processing approach is developed to estimate the frequency of harmonic structural vibrations. This includes a data-aided pipeline for closed-form carrier frequency offset (CFO) estimation and compensation, while residual phase noise is retained in the signal model. The phase uncertainty is described through the Tikhonov distribution. A multi-tier vibration-coherent grid search is then adopted for precise absolute distance estimation. • For multi-target ranging, a frequency-domain dualband processing architecture is investigated. The target response is represented as a finite sum of complex exponentials and recast into an equivalent all-pole linear-prediction model. The proposed estimator combines Hankel covariance processing, Minimum Description Length model-order estimation, and Root-MUSIC initialization. The resulting range estimates are subsequently refined through coherent variable projection and a damped Gauss–Newton procedure. For the SHM application, the results demonstrate reliable vibration-frequency extraction together with high-accuracy distance estimation in the presence of additive noise and residual phase perturbations. For the dual-band multi-target problem, the proposed method achieves millimetre-level mean ranging errors at sufficiently high SNR, while the GOSPA analysis quantifies the degradation associated with low SNR, increasing target cardinality, and reduced per-band spectral support.