Abstract:In beamformed wireless cellular systems such as 5G New Radio (NR) networks, beam management (BM) is a crucial operation. In the second phase of 5G NR standardization, known as 5G-Advanced, which is being vigorously promoted, the key component is the use of artificial intelligence (AI) based on machine learning (ML) techniques. AI/ML for BM is selected as a representative use case. This article provides an overview of the AI/ML for BM in 5G-Advanced. The legacy non-AI and prime AI-enabled BM frameworks are first introduced and compared. Then, the main scope of AI/ML for BM is presented, including improving accuracy, reducing overhead and latency. Finally, the key challenges and open issues in the standardization of AI/ML for BM are discussed, especially the design of new protocols for AI-enabled BM. This article provides a guideline for the study of AI/ML-based BM standardization.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is the challenges faced by Beam Management (BM) in 5G - Advanced networks, especially in high - frequency millimeter - wave (mmWave) communications. Specifically, the paper focuses on how to use artificial intelligence (AI) and machine learning (ML) techniques to improve the performance of beam management. The main objectives include: 1. **Improve accuracy**: Significantly improve the accuracy of beam prediction through AI/ML methods without incurring additional overhead. 2. **Reduce overhead**: Traditional methods bring a large amount of measurement/reference signal (RS) overhead and power consumption when performing beam scanning, which is particularly disadvantageous for mobile scenarios. Using AI/ML for sparse beam scanning and beam prediction can save signaling overhead and provide a large latency advantage. 3. **Reduce latency**: When user equipment (UE) and base stations (gNB) have a large number of beams, sequentially performing the beam management process will result in large measurement and reporting latencies. If an AI/ML solution is adopted, the best beam pair can be predicted from a subset of channel or RSRP measurements, thereby reducing the latency of beam acquisition or tracking without sacrificing prediction accuracy. 4. **Other optimizations**: Beam management is closely related to resource allocation. AI/ML can optimize the allocation of frequency channels and time slots to maximize network efficiency. In addition, AI/ML algorithms can predict and mitigate interference between overlapping beams, improving overall signal quality and network performance. AI can detect and diagnose problems in beam management, such as misalignment or signal degradation, and respond quickly to maintain service quality. AI/ML can also optimize beam management operations, reduce the power consumption of network devices, and contribute to the realization of greener wireless networks. To achieve these goals, the paper discusses the traditional non - AI beam management framework and the AI - based beam management framework, and details two main AI/ML beam management frameworks, namely Spatial Beam Prediction (SBP) and Temporal Beam Prediction (TBP). At the same time, the paper also explores the key challenges and open issues faced in the standardization process, especially the design of new protocols to support AI/ML beam management.

AI/ML for Beam Management in 5G-Advanced: A Standardization Perspective

5G-Advanced AI/ML Beam Management: Performance Evaluation with Integrated ML Models

Going Beyond RF: How AI-enabled Multimodal Beamforming will Shape the NextG Standard

Going beyond RF: A survey on how AI-enabled multimodal beamforming will Shape the NextG standard

Beam Management in 5G: A Stochastic Geometry Analysis

A Survey of Beam Management for mmWave and THz Communications Towards 6G

An Overview of the 3GPP Study on Artificial Intelligence for 5G New Radio

Site-Specific Beam Alignment in 6G via Deep Learning

Machine Learning (ML)-assisted Beam Management in millimeter (mm)Wave Distributed Multiple Input Multiple Output (D-MIMO) systems

Compressed Beam Selection for Single/multi-cell Beam Management

Beam Management with Orientation and RSRP using Deep Learning for Beyond 5G Systems

Efficient Initial Access with Deep Reinforcement Learning Based Beam Sweeping in Wireless Cellular Communication Systems

A Low-Complexity Machine Learning Design for mmWave Beam Prediction

A Deep Learning Framework for Adaptive Beamforming in Massive MIMO Millimeter Wave 5G Multicellular Networks

Deep Learning-Based Mmwave Beam Selection for 5G NR/6G with Sub-6 GHz Channel Information: Algorithms and Prototype Validation

Beam Management Driven by Radio Environment Maps in O-RAN Architecture

Deep Learning for Beam-Management: State-of-the-Art, Opportunities and Challenges

Vision-Assisted mmWave Beam Management for Next-Generation Wireless Systems: Concepts, Solutions and Open Challenges

Artificial Intelligence Enabled Wireless Networking for 5G and Beyond: Recent Advances and Future Challenges

Modular and High-Resolution Channel State Information and Beam Management for 5G New Radio