@article{gupta_bayesian_2025,

title = {Bayesian Learning Aided Parameter Estimation and Joint Beamformer Design in mmWave MIMO-OFDM ISAC Systems},

author = {Awadhesh Gupta and Jitendra Singh and Suraj Srivastava and Aditya K. Jagannatham and Lajos Hanzo},

url = {https://ieeexplore.ieee.org/document/11030617},

doi = {10.1109/TCOMM.2025.3578813},

issn = {1558-0857},

year  = {2025},

date = {2025-01-01},

urldate = {2025-10-08},

journal = {IEEE Transactions on Communications},

pages = {1–1},

abstract = {A three-dimensional (3D) sparse signal recovery problem formulation is conceived for delay, Doppler, and angular (DDA) domain target parameter estimation in millimeter wave (mmWave) multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM)-based integrated sensing and communication (ISAC) systems relying on a hybrid beamforming architecture. Subsequently, a 3D-sparse Bayesian learning (3D-BL) algorithm is proposed to jointly estimate the angular, range, velocity, and radar cross-section (RCS) parameters of the targets. Furthermore, an uplink beamformer is designed for the user equipment (UE) to alleviate the complexity of uplink parameter estimation at the dual-functional radar-communication (DFRC) base station (BS) by eliminating the need for angle of departure (AoD) estimation. Additionally, a Bayesian alternating minimization (BAT-MIN) algorithm is constructed for the designing of a DFRC waveform, enabling the simultaneous generation of beams toward both the radar targets and the UE. Furthermore, the sparse Bayesian learning lower bound (SBL-LB) and the Bayesian Cramér-Rao lower bound (BCRLB) are derived to serve as benchmarks for estimation performance. Finally, simulation results are presented to showcase the enhanced performance of the proposed methodologies in terms of multiple performance metrics when contrasted both to the existing sparse recovery techniques and to conventional non-sparse parameter estimation algorithms. The simulation outcomes unequivocally demonstrate the commendable performance of the proposed 3D-BL estimation methodology, approaching closely to the SBL-LB. Notably, this approach exhibits a substantial gain of at least 5 dB when compared to alternative techniques. Additionally, the introduced BAT-MIN beamformer emerges as a highly competitive solution, closely approximating the capabilities of a fully digital beamformer while maintaining a noteworthy minimum advantage over its contemporaries. These findings underscore the significance and efficacy of the proposed techniques in the context of advanced signal processing and beamforming.},

keywords = {Array signal processing, Bayes methods, dual-functional radar-communication (DFRC), Estimation, hybrid analog-digital (HAD) beamforming, Integrated sensing and communication (ISAC), millimeter wave (mmWave), Millimeter wave communication, MIMO, OFDM, Parameter estimation, Radar, Radar cross-sections, Radio frequency, sparse Bayesian learning (SBL)},

pubstate = {published},

tppubtype = {article}

}