This paper presents a blockage prediction system based on artificial intelligence/machine learning (AI/ML) for millimeter wave (mmWave) communications, with a particular focus on vehicular scenarios. Although numerous related works on ML-based blockage prediction, these works focus on simple scenarios, where the trajectory of the moving object is either fixed or deterministic. In order to make the blockage prediction system more practical, this paper assumes a nondeterministic situation such as both communication terminals and blockers are moving. Furthermore, we have formulated a practical AI/ML-based blockage prediction procedure that is operable in 3GPP compliant systems with taking its beam measurement reports and frame structure. The proposed prediction method uses the measurement report for beam management as input of a neural network. The computer simulation is conducted to show prediction performance of proposed method.

In this paper, we present a novel approach to analog beam selection in cell-free massive multi-input multi-output (CF-mMIMO) systems employing hybrid beamforming (BF) over millimeter-wave (mmWave) channels. In conventional CF-mMIMO systems with hybrid BF, an analog beam is selected by a central processing unit (CPU) from a predefined codebook. This selection is based on the received power at each access point (AP) and the fairness among user equipment (UE). However, this conventional method disregards the potential impact of interference and heavily relies on digital BF to mitigate it. Consequently, when utilizing low-complexity digital BF techniques such as maximum ratio transmission (MRT), the system's performance experiences degradation. Our proposed beam selection method takes into account both the received power and the interference power, effectively mitigating the influence of interference while maintaining an adequate level of received power. The numerical simulations validate the efficacy of our proposed approach.

This paper proposes a clustering scheme suitable for a centralized millimeter wave Cell-Free massive MIMO with hybrid beamforming (BF). Conventionally AP-wise clustering schemes have been proposed for a decentralized architecture, that suppresses multi-user interference at each access point (AP) with local digital precoding and analog BF of each radio frequency (RF) chain, in order to obtain better performance efficiently. However in the centralized architecture, designing the precoder and cluster at central processing unit (CPU) with considering all RF chains of all APs together, the AP-wise clustering could not fit well especially when a large number of UEs are multiplexed. Because the RF chains with high coupling loss (the sum of the path loss and BF gain of the analog beam) could be included in the APs of the cluster to each UE. The proposed method forms a cluster on a per-RF-chain basis, and can select RF chains with lower coupling loss for each UE at the CPU regardless of the AP by incorporating the analog BF gain of each RF chain to form a cluster for each UE. Through simulation evaluations, we show that the hybrid BF with the proposed clustering scheme can obtain the superior spectral efficiency while effectively reducing the complexity of the centralized digital precoding.