Received June 18, 2021, accepted July 1, 2021, date of publication July 6, 2021, date of current version July 13, 2021. Digital Object Identifier 10.1109/ACCESS.2021.3095121 Channel Estimation for 6G V2X Hybrid Systems Using Multi-Vehicular Learning MAROUAN MIZMIZI 1 , (Member, IEEE), DARIO TAGLIAFERRI 1 , (Member, IEEE), DAMIANO BADINI 2 , CHRISTIAN MAZZUCCO 2 , AND UMBERTO SPAGNOLINI 1 , (Senior Member, IEEE) 1 Dipartimento di Elettronica, Informazione e Bioingegneria, Politecnico di Milano, 20133 Milan, Italy 2 Huawei Technologies Italia S.r.l., 20090 Segrate, Italy Corresponding author: Marouan Mizmizi (marouan.mizmizi@polimi.it) This work was supported by the Huawei-Politecnico di Milano Joint Research Laboratory. ABSTRACT Channel estimation for hybrid Multiple Input Multiple Output (MIMO) systems at Millimeter- Waves/sub-THz is a fundamental, despite challenging, prerequisite for an efficient design of hybrid MIMO precoding/combining. Most works propose sequential search algorithms, e.g., Compressive Sensing (CS), that are most suited to static channels and consequently cannot apply to highly dynamic scenarios such as Vehicle-to-Everything (V2X). To address the latter ones, we leverage recurrent vehicle passages to design a novel Multi Vehicular (MV) hybrid MIMO channel estimation suited for Vehicle-to-Infrastructure (V2I) and Vehicle-to-Network (V2N) systems. Our approach derives the analog precoder/combiner through a MV beam alignment procedure. For the digital precoder/combiner, we adapt the Low-Rank (LR) channel estimation method to learn the position-dependent eigenmodes of the received digital signal (after beamforming), which is used to estimate the compressed channel in the communication phase. Extensive numerical simulations, obtained with ray-tracing channel data and realistic vehicle trajectories, demonstrate the benefits of our solution in terms of both achievable spectral efficiency and mean square error compared to the unconstrained maximum likelihood estimate of the compressed digital channel, making it suitable for both 5G and future 6G systems. Most notably, in some scenarios, we obtain the performance of the optimal fully digital systems. INDEX TERMS Low-rank channel estimation, hybrid MIMO systems, millimeter-wave, sub-THz, V2X, 5G new radio, 6G. I. INTRODUCTION Recent advances in millimeter-wave (mmW) hardware [1] and the potential availability of spectrum has encouraged the wireless industry to consider mmW, for the Fifth Gen- eration of cellular systems (5G) [2] and, in particular, for Vehicle-to-Everything (V2X) applications [3], [4]. Following the same trend, sub-THz are envisioned for 6G systems [5]–[7]. Due to the increased carrier frequency, e.g., 24.25 − 52.6 GHz for 5G New Radio (NR) Frequency Range 2 (FR2) and >100 GHz for sub-THz, mmW/sub-THz signals experience an orders-of-magnitude increase in free-space path loss compared to the current majority of wireless systems, resulting in highly sparse channels [8], [9]. Multiple Input Multiple Output (MIMO) systems are a redeeming solution that can provide a beamforming The associate editor coordinating the review of this manuscript and approving it for publication was Mohammad S. Khan . gain to overcome the path loss and establish links with a rea- sonable Signal-to-Noise Ratio (SNR). Additionally, MIMO systems enable precoding and combining of multiple data streams which could significantly improve the achievable data rate [10], [11]. While the fundamental theory of MIMO precod- ing/combining is the same regardless of the carrier frequency, the hardware in the mmW/sub-THz band is subject to a set of non-trivial practical limitations. The processing in traditional MIMO systems is performed digitally at baseband, which requires a dedicated Radio Frequency (RF) chain for each antenna element. Unfortunately, due to the high number of elements required in mmW (even more at sub-THz), this implies a high cost and power consumption, which makes it unpractical [12]. A promising solution to these problems lies in the con- cept of hybrid arrays, which use a combination of analog beamforming in the RF domain and digital beamforming VOLUME 9, 2021 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ 95775