Near-Field Terahertz Communications: Model-Based and Model-Free Channel Estimation

Terahertz (THz) band is expected to be one of the key enabling technologies sixth generation (6G) wireless networks because its abundant available bandwidth and very narrow beamwidth. Due high frequency operations, electrically small array apertures are employed, signal wavefront becomes spherical in near-field. Therefore, near-field model should considered for channel acquisition THz systems. Unlike prior works which mostly ignore impact beam-squint (NB) consider either narrowband scenario or far-field models, this paper introduces both a model-based model-free techniques wideband estimation presence NB. The approach based on orthogonal matching pursuit (OMP) algorithm, we design an NB-aware dictionary. idea exploit angular range deviations due We then employ OMP accounts thereby ipso facto mitigating effect further introduce federated learning (FL)-based as solution multi-user achieve reduced complexity training overhead. Through numerical simulations, demonstrate effectiveness proposed systems comparison with existing state-of-the-art techniques.