Terahertz (THz) band communications, renowned for their ultra-wide bandwidth of several hundred gigahertz (GHz), are pivotal for meeting the growing demands of wireless data traffic in the upcoming sixth-generation (6G) wireless communication era. However, the utilization of extensive bandwidth and large antennas in THz communications introduce challenges, including delay and beam squint effects, collectively termed dual wideband effects. In this work, we explore an iterative atomic norm minimization (ANM)-based gridless frequency-selective channel estimation tailored for THz ultra-massive multiple-input multiple-output orthogonal frequency division multiplexing (UM-MIMO-OFDM) systems under dual wideband effects. Unlike prevailing research focused on on-grid compressed sensing (CS)-based channel estimation, this approach surpasses on-grid CS-based methods such as sparse Bayesian learning (SBL), orthog-onal matching pursuit (OMP), generalized simultaneous orthog-onal matching pursuit (GSOMP), and classical least squares (LS) channel estimators. Its superiority lies in overcoming the grid mismatch problem, a prevalent issue in on-grid CS-based methods, as demonstrated through evaluation using normalized mean square error (NMSE) as a key metric.
Gridless Channel Estimation for THz Ultra-Massive MIMO-OFDM Systems Under Dual Wideband Effects
24.06.2024
515840 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch