PaperlayerView exact revision on arXiv

Source-linked AI summary

6G: The Intelligent Network of Everything

Harri Pennanen, Tuomo Hänninen, Oskari Tervo, Antti Tölli, Matti Latva-aho

arXiv:2407.09398v3cs.ITmath.IT

TL;DR

This article offers a broad guide to 6G, bringing together its vision, technology survey, and tutorial coverage. It presents 6G as the Intelligent Network of Everything and places that vision in a wider perspective by sketching what may follow 6G.

  • Problem

    The article aims to provide a comprehensive guide that brings together a 6G vision, literature survey, and tutorial treatment of potential technologies.

    Source: §XIII · p. 84

  • Method

    The article surveys 6G visions and technologies, explains potential technologies in tutorial style, and sketches a high-level vision of 7G.

    Source: §XIII · p. 84

  • Results

    The article presents 6G as the Intelligent Network of Everything and frames it as a platform for mobile intelligence.

    Source: §III · p. 17, §XIII · p. 84

  • Takeaways & Limitations

    The article places its 6G vision in a wider perspective by introducing a high-level vision of 7G.

    Source: §XIII · p. 84

Abstract

from arXiv · show

The global 6G vision has taken its shape after years of international research and development efforts. This work culminated in ITU-R Recommendation on "IMT-2030 Framework". While the definition phase of technological requirements is currently ongoing, 3GPP standardization process on 6G networks is expected to start in 2025 and worldwide commercialization around 2029-2030. This article serves as a comprehensive guide to 6G by providing an overall vision, a contemporary survey of the main literature, and an informative tutorial-type presentation style. In our vision, 6G will be based on three fundamental elements: wireless, artificial intelligence, and Internet of Everything. Consequently, 6G can ultimately become the Intelligent Network of Everything while serving as an enabling platform for the next major disruption in mobile communication, called mobile intelligence. The potential of mobile intelligence is that anything can be made connected, intelligent, and aware of its environment. This will revolutionize the way how devices, systems, and applications are designed; how they operate and interact with humans and each other; and how they can be used for the benefit of people, society, and the world in general. After high-level visioning, the main details of 6G are discussed, including fundamental elements, disruptive applications, key use cases, main performance requirements, potential technologies, and defining features. A special focus is given to a comprehensive set of potential 6G technologies, each of which is introduced in a tutorial manner. Finally, we speculate on what comes after 6G and sketch the first high-level vision of 7G. All in all, the objective of this article is to provide a thorough guide to 6G in order to serve as a source of knowledge and inspiration for further research and development work in academia, industry, and standardization bodies.

I. INTRODUCTION

The article traces mobile communications from 1G to 6G, reviews 5G and its standardization, and introduces 6G as mobile intelligence. It also surveys 6G development and research activity to situate the article’s broader vision.

I. INTRODUCTION

  • The article surveys 6G’s development, research activity, and prior vision work before presenting its own vision and technology tutorial.Its introduction also previews a discussion of what may follow 6G.Source: §I · p. 1

1) 1G: MOBILE MEETS TELEPHONY

4) 4G: MOBILE INTERNET

  • 4G’s broadband internet, OFDM, and MIMO helped mobile internet become an everyday service.LTE and its evolution added technologies including carrier aggregation, enhanced MIMO, and V2X and MTC support.Source: 4) 4G: MOBILE INTERNET · p. 3

6) RELEASE 18: 5G-ADVANCED

6) 6G: MOBILE INTELLIGENCE

  • 6G is presented as the Intelligent Network of Everything, combining wireless, AI/ML, and IoE with beyond-communication capabilities.The article associates this vision with mobile intelligence and connected, intelligent, environment-aware devices, systems, and applications.Source: 6) 6G: MOBILE INTELLIGENCE · p. 3

B. 5G IN A NUTSHELL

  • 5G supports eMBB, URLLC, and mMTC, with requirements including 20 Gbit/s peak rate and 1 ms latency.The 5G System is described as the dominant technology used worldwide.Source: B. 5G IN A NUTSHELL · p. 4

2) BACKGROUND

  • The 5G standardization path includes ITU-R evaluation of candidate radio interfaces, with 5G NSA relying on 4G LTE and 5G SA operating independently.The article describes the ITU-R evaluation as culminating in the 2021 IMT-2020 Recommendation.Source: 2) BACKGROUND · p. 4

7) RELEASE 19: 5G-ADVANCED EVOLUTION

The article describes Release 19 as preparation for 6G, then outlines the roles and timelines of ITU-R and 3GPP. It closes this section with a chronological survey of worldwide 6G research activity from 2018 through 2021.

7) RELEASE 19: 5G-ADVANCED EVOLUTION

  • Release 19 continues 5G-Advanced while preparing studies toward 6G, including channel modeling, AI/ML for the air interface, and energy saving.Its stated focus also includes NTN and IoT evolution, ambient IoT, XR, and mobility improvements.Source: 7) RELEASE 19: 5G-ADVANCED EVOLUTION · p. 5

2) ROLE OF ITU-R

  • ITU-R’s three phases cover the IMT-2030 vision, requirements and evaluation methods, then candidate evaluation and selection by 2030.The requirements and methodology phase is expected to run from 2024 to 2026; candidate evaluation is planned for 2027–2030.Source: 2) ROLE OF ITU-R · p. 6
  • ITU-R’s phase 1 produced a technology-trends report and the IMT-2030 vision Recommendation in 2022 and 2023.The vision document covers trends, usage scenarios, capabilities, and ongoing development.Source: 2) ROLE OF ITU-R · p. 6
  • ITU-R identified mid-band ranges for IMT-2030 studies and above-100-GHz ranges for preliminary studies toward possible later approval.The mid-band ranges are described as favorable for wide coverage; higher bands are associated with capacity and ultra-broadband links in dense hotspots.Source: 2) ROLE OF ITU-R · p. 6

3) ROLE OF 3GPP

  • 3GPP develops mobile standards through releases, with study items exploring feasibility and work items producing specifications intended for formal standards.Release 20 is expected to study 6G features, while Release 21 is expected to define 6G technologies.Source: 3) ROLE OF 3GPP · p. 7

D. 6G RESEARCH ACTIVITIES

  • The article’s research-activity survey is chronological and spans initiatives, projects, collaborations, funding, experiments, events, and publications.The authors state that their identification of main activities and benchmarks is based on an extensive 6G literature search.Source: D. 6G RESEARCH ACTIVITIES · p. 7

1) ACTIVITIES IN 2018

  • In 2018, the survey identifies Finland’s 6G Flagship as the first large-scale 6G project, alongside early vision papers and a workshop series.The project’s reported volume was 249 million euros and its aim was to develop and test 6G enabling technologies.Source: 1) ACTIVITIES IN 2018 · p. 7

2) ACTIVITIES IN 2019

  • From 2019 to 2021, 6G activity spread through international projects, partnerships, trials, white papers, conferences, and research programs.Reported examples include the 2020 IMT-2030 visioning work and 2021 demonstrations of THz links and related technologies.Source: 2) ACTIVITIES IN 2019 · p. 8, 3) ACTIVITIES IN 2020 · p. 8

5) ACTIVITIES IN 2022

The article chronicles worldwide 6G activity from 2022 through 2024, alongside a broad vision, survey, and tutorial spanning 6G technologies, applications, requirements, and future generations. Its contributions include a three-pillar 6G vision and a high-level sketch of 7G.

5) ACTIVITIES IN 2022

  • In 2022, 6G work spanned prototypes and trials, spectrum and use-case papers, international partnerships, research facilities, and the first IMT-2030 trends report.Examples include a 320-meter outdoor THz test and publication of the first IMT-2030 vision recommendation.Source: 5) ACTIVITIES IN 2022 · pp. 10–11

6) ACTIVITIES IN 2023

  • In 2023, activity included new research projects and collaborations, sub-THz demonstrations, and ITU-R’s first IMT-2030 Framework recommendation.Reported demonstrations included 25 Gbit/s at 144 GHz and over 100 Gbit/s at a 300 GHz carrier frequency.Source: 6) ACTIVITIES IN 2023 · pp. 11–12

7) ACTIVITIES IN 2024

  • ITU-R began the second phase of IMT-2030 development, defining its requirements and corresponding evaluation methodologies; other reported 6G work included digital-twin efforts, NTN work, partnerships, and an upper-6-GHz field trial.Source: 7) ACTIVITIES IN 2024 · pp. 12–13

E. CONTRIBUTIONS OF THE ARTICLE

  • The article packages a 6G vision, literature survey, and tutorial, including reviews of 27 potential technologies across 10 categories.Each technology uses a tutorial template covering vision, background, opportunities and challenges, and literature and future directions.Source: E. CONTRIBUTIONS OF THE ARTICLE · p. 13
  • The article organizes 12 disruptive applications into three categories and identifies five communication-oriented and three beyond-communication-oriented use cases.It also reviews AI/ML concepts, promising methods, and integration opportunities and challenges.Source: E. CONTRIBUTIONS OF THE ARTICLE · p. 13
  • The article compares IMT-2030 requirements with typical literature values and surveys 6G development, worldwide research activity, prior visions and surveys, and 5G.The review of 6G research activity covers projects, collaborations, funding, standardization, experiments, events, and publications; the prior-work review comprises over 60 articles.Source: E. CONTRIBUTIONS OF THE ARTICLE · pp. 13–15
  • Its vision grounds 6G in wireless, AI, and IoE, and identifies 12 defining features, including intelligence, sustainability, reliability, and security.The article describes 6G as an Intelligent Network of Everything serving as a platform for mobile intelligence.Source: E. CONTRIBUTIONS OF THE ARTICLE · p. 13
  • The article discusses what may follow 6G, sketches a high-level vision for 7G, and describes an evolution path from 1G to 7G with cycles of two generations.The 7G vision focuses on the main disruption, fundamental elements, and possible applications of 7G.Source: E. CONTRIBUTIONS OF THE ARTICLE · p. 13

II. PREVIOUS WORKS

Prior 6G work spans visions and surveys of the field, while the authors identify a need for a current, broad guide and present a vision extending from mobile intelligence to the Intelligent Network of Everything.

A. 6G VISION ARTICLES

  • Early 6G vision papers covered interdisciplinary research, usage scenarios, target requirements, AI-based network design, and promising technologies, as well as AI-empowered network architecture and technologies.Source: A. 6G VISION ARTICLES · p. 15
  • Later vision work addressed AI/ML, architecture, requirements, applications, and technology enablers, including European 6G initiatives and specific use-case visions,,,,.The reviewed topics also include broadband connectivity, societal change, and the evolution of mobile networks.Source: A. 6G VISION ARTICLES · pp. 15–16

B. 6G SURVEY ARTICLES

  • 6G surveys collectively review network requirements, applications, architectures, technologies, open challenges, and research progress,,,,,.Their scopes range from particular facets such as IoT and positioning to end-to-end system designs.Source: B. 6G SURVEY ARTICLES · pp. 16–17

C. DISCUSSION

  • The authors identify three recurring themes—improved performance, more connected objects, and a key role for AI/ML—and say the literature lacks an up-to-date overview of the whole field.They present their contemporary survey and tutorial, alongside a high-level vision beyond 6G, as addressing that gap.Source: C. DISCUSSION · p. 17

III. 6G VISION

  • The article introduces a 6G vision aligned with ITU-R but described as broader, centered on mobile intelligence and an Intelligent Network of Everything.Mobile intelligence is framed as wireless, AI, and IoE fused across society, with the potential to make things connected, smart, and environment-aware.Source: §III · p. 17
  • The vision rests on wireless, AI, and IoE: communication, sensing, and energy; AI across network layers and applications; and many connected objects.Figure 7 illustrates the three fundamental elements.Source: §III · p. 17, Figure 7 · p. 19
  • The article groups disruptive applications into human-machine interactions, smart environments, and connected autonomous systems, and identifies five communication-oriented and three beyond-communication use cases.Figure 6 illustrates the smart-wireless-world vision; the application groups entail different network needs, including low latency, customized services, and reliability.Source: Figure 6 · p. 17, §III · p. 18
  • ITU-R example targets span nine performance metrics and are described as moderate against literature proposals; the article reviews ten technology categories and names six grouped defining-feature families.The feature groups include intelligence and awareness, availability and reliability, sustainability, and security; the technology review covers areas such as spectrum, network intelligence, energy awareness, and security.Source: §III · pp. 18–19

IV. FUNDAMENTAL ELEMENTS OF 6G

The section presents wireless, AI, and the Internet of Everything as 6G’s foundational elements, then surveys applications and smart environments that they may support. It covers human-machine interactions from metaverse to Tactile Internet, alongside society, city, factory, and home settings.

IV. FUNDAMENTAL ELEMENTS OF 6G

  • 6G’s three foundations are wireless technologies, pervasive AI, and the Internet of Everything (IoE), spanning communication and beyond-communication capabilities and diverse connected entities.Wireless includes sensing, positioning, and energy; AI is envisioned across core, edge, and air interface, while IoE extends connectivity beyond sensors to machines, systems, and applications.Source: §IV · p. 19

V. DISRUPTIVE APPLICATIONS FOR 6G

  • The article introduces 12 disruptive 6G applications across human-machine interactions, smart environments, and connected autonomous systems.Source: §V · p. 19

1) METAVERSE

  • The metaverse is envisioned as an immersive, interoperable virtual ecosystem merging physical, human, and digital worlds; its proposed evolution runs from digital twins to digital natives and surreality.The passage describes the metaverse as still in its infancy and 6G as a potential platform for its development.Source: 1) METAVERSE · pp. 19–20

2) EXTENDED REALITY

  • XR combines real and virtual environments, but current headsets’ cost, weight, and wiring limit uptake, while future immersive applications impose demands 5G cannot meet.The envisioned XR experience adds haptics to audiovisual content, with smell and taste as longer-term possibilities.Source: 2) EXTENDED REALITY · p. 20

3) HOLOGRAPHIC-TYPE COMMUNICATION

  • Holographic communication transfers view-dependent 3D representations; high-quality light-field displays can require Tbps-level rates and sub-millisecond latency.The paper describes holograms as generated or stored at a source, transferred over a network, and presented on a display.Source: 3) HOLOGRAPHIC-TYPE COMMUNICATION · pp. 20–21

4) DIGITAL TWINS

  • Digital twins connect a physical object or entity to its virtual copy, enabling monitoring, analysis, evaluation, control, and management using collected data.The paper lists applications including manufacturing, network management, healthcare, and smart cities.Source: 4) DIGITAL TWINS · p. 21

5) TACTILE INTERNET

  • Tactile Internet adds haptic information for real-time human-machine control and interactivity, requiring latency up to 1 ms and reliability up to 1-10-7, depending on the application.The passage says 5G can support light versions but was not designed to enable demanding Tactile Internet applications in practice.Source: 5) TACTILE INTERNET · p. 21

B. SMART ENVIRONMENTS

  • Smart environments apply digital technologies to create value across society, cities, factories, and homes; smart factories target greater efficiency, while smart-home solutions remain fragmented and incomplete.The passage says smart-city priorities depend on local factors, and identifies inadequate common policies and data security and privacy standardization as smart-home gaps.Source: B. SMART ENVIRONMENTS · p. 22, A smart factory · p. 22, A smart home · p. 22

C. CONNECTED AUTONOMOUS SYSTEMS

The section surveys connected autonomous vehicles, aerial vehicles, and robots, then organizes 6G use cases around communication and beyond-communication capabilities. It also reviews ITU-R’s IMT-2030 performance framework and selected targets.

1) CONNECTED AUTONOMOUS VEHICLE SYSTEMS

  • Connected autonomous vehicle systems combine autonomous driving with V2X connectivity; the paper says 6G is intended to better support their ubiquitous, reliable connectivity needs.Vehicle autonomy is divided into six SAE levels, from no automation to full automation.Source: 1) CONNECTED AUTONOMOUS VEHICLE SYSTEMS · p. 23

2) CONNECTED AUTONOMOUS AERIAL VEHICLE SYSTEMS

3) CONNECTED AUTONOMOUS ROBOTIC SYSTEMS

  • Connected autonomous robotic systems use AI/ML-based control and network connectivity to coordinate independent robots, with potential applications across manufacturing, logistics, healthcare, and other sectors.The paper identifies secure, high-performance connectivity, minimal delays, reliability, and IoT capabilities as potential 6G support.Source: 3) CONNECTED AUTONOMOUS ROBOTIC SYSTEMS · pp. 23–24

A. COMMUNICATION-ORIENTED USE CASES

  • 6G’s five communication-oriented use cases address capacity, latency, reliability, density, coverage, and mobility through broadband, mission-critical, massive, global-scale, and hyper-mobility communications.The listed use cases include support for high-velocity communications up to 500–1000 km/h.Source: A. COMMUNICATION-ORIENTED USE CASES · pp. 24–25

B. BEYOND-COMMUNICATION-ORIENTED USE CASES

  • The beyond-communication use cases are network intelligence, network sensing, and network energy, extending cellular capabilities to AI/ML services, sensing and positioning, and wireless energy transfer.The cited table summarizes 6G use cases; the prose identifies ubiquitous communication, computation, and caching infrastructure as a challenge for network intelligence.Source: B. BEYOND-COMMUNICATION-ORIENTED USE… · pp. 25–26, Table 4 · p. 26

VII. MAIN PERFORMANCE REQUIREMENTS FOR 6G

  • ITU-R’s IMT-2030 Framework identifies 15 performance capabilities and numerical targets for nine; the article compares these with IMT-2020 and reviews literature-proposed values.The passage notes that ITU-R leaves some capabilities without numerical targets because they are difficult to measure numerically.Source: §VII · p. 26
  • For IMT-2030, ITU-R target peak data rates are 50, 100, and 200 Gbit/s by usage scenario, while latency targets range from 0.1 to 1 ms.The passage separately reports literature targets such as 1 Tbit/s peak rate and 0.1 ms latency.Source: §VII · pp. 26–27
  • ITU-R proposes IMT-2030 connection densities of 106–108/km2 and mobility of 500–1000 km/h; it suggests positioning accuracy of 1 to 10 cm.Source: §VII · p. 27

VIII. AI/ML FOR 6G

This section introduces AI/ML fundamentals and learning modes, then surveys deep, federated, and transfer learning as methods with potential benefits and challenges for 6G.

A. INTRODUCTION TO AI AND ML

  • AI describes computer programs performing tasks regarded as intelligent, while ML learns task performance from training data and is useful when mathematical models or algorithms are lacking.The section also notes that some multi-phase problems may benefit from combining model-based and data-based approaches.Source: 1) ARTIFICIAL INTELLIGENCE · p. 27, 2) MACHINE LEARNING · p. 27

3) MAIN TYPES OF ML

  • Supervised learning uses labeled input-output pairs for tasks such as detection and prediction, while unsupervised learning finds patterns in unlabeled data for tasks such as compression and clustering.Reinforcement learning uses feedback from actions and is suited to sequential decisions, including communication resource allocation.Source: 3) MAIN TYPES OF ML · p. 28

4) MAIN TYPES OF TRAINING

  • Offline learning trains over the full dataset before deployment, whereas online learning updates incrementally and can adapt to changing conditions with fewer computation and storage resources.Offline retraining can be costly and poorly scalable; online learning may face efficiency and robustness issues when training data are poor.Source: 4) MAIN TYPES OF TRAINING · p. 28

B. KEY ML METHODS FOR 6G

1) DEEP LEARNING

  • DL uses multilayer neural networks to transform raw inputs into higher-level representations and can be accurate and efficient when training datasets are sufficiently large.It applies under supervised, unsupervised, or reinforcement-learning principles and spans architectures such as convolutional and recurrent neural networks.Source: 1) DEEP LEARNING · pp. 28–29

2) FEDERATED LEARNING

  • FL trains models locally on edge devices and exchanges model parameters rather than raw data, potentially saving wireless resources, reducing latency, and enhancing privacy.Latency reductions depend on convergence speed, and practical deployment still faces challenges including communication overhead, scalability, heterogeneity, and security.Source: 2) FEDERATED LEARNING · p. 29

3) TRANSFER LEARNING

  • TL transfers knowledge from source to target domains to reduce training-data needs and computational demands, but dissimilar sources can cause negative transfer.The section identifies advanced forms including deep, federated, and online TL as relevant to 6G.Source: 3) TRANSFER LEARNING · pp. 29–30

C. DISCUSSION

The discussion surveys candidate 6G technologies and spectrum options while highlighting technical, energy, privacy, and reliability challenges. It also outlines OWC research categories and priorities, emphasizing interdisciplinary cooperation to advance 6G-ready systems.

C. DISCUSSION

  • AI/ML could reshape 6G network design, operation, and management, but deployment requires robust computing from core to edge, data processes, real-time processing, and energy efficiency.Extensive collection of potentially sensitive data raises privacy, ownership, and consent concerns; limited transparency also complicates accountability, motivating clear regulation and responsible AI practices.Source: C. DISCUSSION · p. 30

IX. POTENTIAL TECHNOLOGIES FOR 6G

  • 6G technology development surveys 27 potential technologies across 10 categories to support the complexity and heterogeneity of future networks.The categories and technologies are introduced in Figure 11.Source: §IX · p. 30

A. SPECTRUM-LEVEL TECHNOLOGIES FOR 6G

  • 6G must support spectrum from sub-6 GHz and cmWave through mmWave and THz, balancing lower bands' coverage and mobility against higher bands' capacity and throughput.Optical wireless may add complementary spectrum for short-range indoor access and long-range outdoor backhaul.Source: A. SPECTRUM-LEVEL TECHNOLOGIES FOR 6G · p. 31

1) THZ COMMUNICATIONS •

  • 100+ Gbit/s at 300 GHz was demonstrated by Keysight and partners in 2023, while Nokia and collaborators demonstrated 25 Gbit/s at 144 GHz.Other reported demonstrations included Samsung's 6.2 Gbit/s at 140 GHz over 50 feet in 2021 and LG's 320-meter outdoor link at 155–175 GHz in 2022.Source: 1) THZ COMMUNICATIONS • · p. 32
  • THz's massive bandwidth can support 1 Tbit/s peak rates, while narrow beams and large bandwidths also enable sensing, localization, and reduced latency.THz also offers close-proximity connectivity and potential PHY-layer security, but short range from high propagation losses is a central limitation.Source: 1) THZ COMMUNICATIONS • · pp. 31–32
  • THz integration faces high propagation loss, blockages, hardware impairments, difficult transceiver design, high power use, and beam misalignment.Ultra-dense deployments and ultra-massive MIMO beamforming may address short range, but accurate beamforming remains difficult in practice.Source: 1) THZ COMMUNICATIONS • · p. 32

2) OPTICAL WIRELESS COMMUNICATIONS •

  • Optical wireless can complement 6G with short-range indoor access and long-range outdoor backhaul, while VLC and FSO offer distinct short- and long-range roles.Experiments have achieved multi-Tbit/s FSO over multiple kilometers and Petabit/s rates over multiple meters, but commercial terrestrial FSO use remains limited by poor link reliability.Source: 2) OPTICAL WIRELESS COMMUNICATIONS • · pp. 34–35
  • OWC's benefits include large bandwidth, low energy use, low latency, RF-interference immunity, spatial reuse, and secure links, but VLC and FSO face distinct reliability challenges.VLC is affected by limited range, blockages, ambient light, line-of-sight needs, receiver orientation, and mobility; FSO reliability is impaired by weather, beam divergence, and pointing errors.Source: 2) OPTICAL WIRELESS COMMUNICATIONS • · p. 35
  • OWC fundamentals require continued development of more spectrum-, energy-, and cost-efficient devices, transceivers, and PHY-layer techniques.The discussion frames these as core elements of OWC systems and calls for cooperation among academia, industry, and standardization bodies.Source: 2) OPTICAL WIRELESS COMMUNICATIONS • · p. 36
  • Future OWC work should develop realistic 6G-scenario channel models and advance network-level design, particularly hybrid RF-optical networks for VLC and FSO.This extends research beyond individual links and PHY techniques to network management and higher-layer considerations; channel modeling is discussed in – and hybrid networks in –.Source: 2) OPTICAL WIRELESS COMMUNICATIONS • · p. 36
  • AI/ML and IRSs are emerging OWC research directions, with AI/ML spanning communication layers and optical IRSs potentially relaxing FSO's line-of-sight requirement.AI/ML is especially relevant to hybrid RF-optical networks; IRSs may also extend coverage, avoid obstacles, and support energy- and cost-efficient deployments.Source: 2) OPTICAL WIRELESS COMMUNICATIONS • · p. 36
  • 6G-relevant OWC applications range from indoor access and outdoor backhauling to D2D/V2V, optical IoT, localization/sensing, and satellite communications.Experiments and prototypes provide benchmarks, and future work should conduct more practical demonstrations tailored to 6G-specific scenarios.Source: 2) OPTICAL WIRELESS COMMUNICATIONS • · p. 37
  • Future VLC and FSO research should address VLC's distance, blockage, uplink, ambient-light, and mobility limits, and improve FSO reliability against pointing errors and atmospheric effects.AI/ML-aided hybrid RF-VLC can switch access to alleviate VLC problems; FSO directions include pointing, acquisition, and tracking, diversity, and hybrid RF-FSO systems for 6G backhaul scenarios,. Human-centric public networks face less predictable blockages than machine-centric private networks; hybrid RF-VLC research is discussed in,.Source: 2) OPTICAL WIRELESS COMMUNICATIONS • · p. 37

B. ANTENNA SYSTEM TECHNOLOGIES FOR 6G

6G antenna systems are expected to extend massive MIMO toward ultra-massive arrays and programmable wireless environments through RISs. Their potential gains are substantial, but practical deployment still faces hardware, control, channel-estimation, and mobility challenges.

1) ULTRA-MASSIVE MIMO •

  • Hundreds to thousands of antennas can form narrow beams for THz coverage and high spatial multiplexing, but hardware cost, power, complexity, and beam management remain challenges.Hybrid beamforming can support multi-stream communication with fewer RF chains than digital beamforming; narrow beams also complicate access, channel estimation, and mobility support.Source: 1) ULTRA-MASSIVE MIMO • · p. 38

2) RECONFIGURABLE INTELLIGENT SURFACES •

  • A 2018 NTT DOCOMO 28 GHz trial reported 560 Mbit/s with a metasurface reflect-array, compared with 60 Mbit/s without reflector assistance.Later trials and prototypes included transparent dynamic metasurfaces and four RIS types showcased by LG in 2022.Source: 2) RECONFIGURABLE INTELLIGENT SURFACES • · p. 40
  • RISs use electronically controlled elements to shape reflected signals, potentially improving signal quality, coverage, and obstacle avoidance with low-cost, energy-efficient deployment.Passive RISs cannot send or receive pilot signals, making channel estimation, CSI acquisition, real-time control, and mobility management particularly problematic.Source: 2) RECONFIGURABLE INTELLIGENT SURFACES • · pp. 39–40
  • RISs could make an otherwise uncontrollable propagation environment programmable, but large-scale deployment requires better models, hardware, control, channel estimation, and resource allocation.Adding an RIS also complicates resource allocation.Source: 2) RECONFIGURABLE INTELLIGENT SURFACES • · p. 40

3) HOLOGRAPHIC MIMO •

The section surveys holographic MIMO and transmission-scheme technologies proposed for 6G. It highlights their potential gains alongside practical trade-offs in implementation, flexibility, reliability, and receiver complexity.

3) HOLOGRAPHIC MIMO •

  • Active HMIMO surfaces could approach massive MIMO’s theoretical performance with high spatial multiplexing and lower energy and cost, but theory, channel models, algorithms, hardware, and experiments remain immature.Near-field propagation and the very large number of elements add modeling and channel-estimation challenges.Source: 3) HOLOGRAPHIC MIMO • · pp. 41–42

1) MULTI-WAVEFORM SCHEME •

  • A flexible multi-waveform scheme is proposed for 6G’s sub-6 GHz-to-THz range, since no single waveform fits all scenarios.An enhanced OFDM-based primary waveform could be complemented by single-carrier DFT-s-OFDM at THz frequencies, where low PAPR and robustness to hardware impairments are useful.Source: 1) MULTI-WAVEFORM SCHEME • · pp. 43–44

2) ADVANCED MODULATION AND CODING METHODS

  • Modulation and coding must balance throughput against reliability: higher modulation orders and lower coding rates favor throughput, while lower orders and higher rates favor reliability.Polar codes are presented as promising for stringent reliability and latency, while higher-order QAM gains at higher frequencies may be limited by power-amplifier inefficiency.Source: 2) ADVANCED MODULATION AND CODING… · pp. 44–45

3) NON-ORTHOGONAL MULTIPLE ACCESS •

  • NOMA can serve multiple devices on one time-frequency resource, making grant-free NOMA a candidate for massive IoT, but receiver complexity and suitable use cases remain challenges.The motivating scenario has more devices requesting service than orthogonal resource units, with sensors sending small, intermittent uplink data.Source: 3) NON-ORTHOGONAL MULTIPLE ACCESS • · pp. 46–47

4) GRANT-FREE MEDIUM ACCESS •

The section surveys grant-free access and four network architectures proposed for 6G, outlining their benefits, challenges, and research directions. Together, they address fast massive access, coverage, capacity, and flexible deployment.

4) GRANT-FREE MEDIUM ACCESS •

  • Grant-free access can reduce latency, signaling overhead, and energy use for IoT devices sending small packets, but collisions and lack of prioritization remain drawbacks.Massive connectivity, tight latency and reliability requirements, and heterogeneous networks compound these challenges; traffic prediction and collision-avoidance strategies are among the proposed approaches.Source: 4) GRANT-FREE MEDIUM ACCESS • · pp. 47–48

1) INTEGRATED NON-TERRESTRIAL AND TERRESTRIAL NETWORKS

  • Integrated non-terrestrial and terrestrial networks combine ground, air, and space layers to extend coverage and support applications such as global IoT and remote-area connectivity.Satellite, HAPS, and UAV layers offer distinct coverage, cost, and deployment trade-offs, while heterogeneous network management, channel modeling, and regulation remain challenges.Source: 1) INTEGRATED NON-TERRESTRIAL AND… · pp. 49–51

2) ULTRA-DENSE NETWORKS •

  • Ultra-dense networks use many heterogeneous access nodes to meet 6G capacity demands, particularly through dense THz deployments, but interference and mobility complicate management.Frequent handovers consume resources, and conventional fiber backhaul can make large-scale deployment costly and slow.Source: 2) ULTRA-DENSE NETWORKS • · pp. 51–52

3) INTEGRATED ACCESS AND BACKHAUL •

  • Integrated access and backhaul uses the same base-station radio resources for wireless access and backhaul, offering a flexible, lower-cost alternative to fiber as networks densify.Resource allocation between access and backhaul depends on traffic loads, while interference management remains a central challenge.Source: 3) INTEGRATED ACCESS AND BACKHAUL • · pp. 52–53

4) CELL-FREE MASSIVE MIMO •

Cell-free massive MIMO is presented as a complement to cellular networks for more stable coverage, while network intelligence extends AI/ML across the core and edge. The discussion highlights deployment constraints and unresolved practical challenges.

4) CELL-FREE MASSIVE MIMO •

  • Cell-free massive MIMO uses distributed low-cost access points to jointly serve nearby users, aiming for stable coverage-wide QoS and improved energy efficiency.Scalability, synchronization, resource allocation, mobility, and the reliance on theoretical channel models remain challenges; practical validation, experiments, and field trials are still needed.Source: 4) CELL-FREE MASSIVE MIMO • · pp. 53–55

E. NETWORK INTELLIGENCE TECHNOLOGIES FOR 6G

  • 6G network intelligence is envisioned as pervasive AI/ML across the core, edge, and air interface, building on AI/ML adoption in 5G Release 18 onward.The article identifies these network levels as cornerstones of 6G network intelligence.Source: E. NETWORK INTELLIGENCE TECHNOLOGIES… · p. 55

1) INTELLIGENT CORE •

  • The core network manages overall network operation, management, and security; pervasive AI/ML is expected to make the 6G core intelligent.Realizing this vision requires capable infrastructure, efficient and reliable algorithms, data acquisition, core-edge cooperation, and security.Source: 1) INTELLIGENT CORE • · p. 56

2) INTELLIGENT EDGE •

  • Edge intelligence combines edge computing and AI, placing computation nearer information sources to reduce latency and communication overhead.Its six training-and-inference arrangements trade off offloading and information exchange against edge-device computation latency and energy consumption.Source: 2) INTELLIGENT EDGE • · p. 57
  • Large-scale edge intelligence requires capable edge infrastructure, real-time processing, suitable AI/ML approaches, available data, and protections for security and privacy.The source identifies cost-performance trade-offs in infrastructure and calls for further study and development to reach the maturity needed for standardization and commercialization.Source: 2) INTELLIGENT EDGE • · pp. 58–59

AI ENHANCES INDIVIDUAL BLOCKS

This section surveys AI/ML-enhanced air-interface design and 6G integration of communication with computation, sensing, and energy, including wireless energy transfer for IoT. It outlines potential capabilities and sustainability support alongside design, implementation, and operational challenges and related research areas.

AI REPLACES MANY BLOCKS JOINTLY

  • AI/ML can advance the air interface from enhancing individual processing blocks to jointly replacing multiple blocks and ultimately learning PHY/MAC functions end to end.Conventional PHY modules can perform well individually yet be suboptimal as a whole, while joint design is mathematically complex—motivating AI/ML approaches [79],.Source: 3) INTELLIGENT AIR INTERFACE • · p. 59, Figure 32 · p. 59

3) INTELLIGENT AIR INTERFACE •

  • AI/ML can improve air-interface performance and simplify design and standardization, but pervasive deployment must address 6G’s complexity and stringent requirements.Challenges include changing channels and environments, diverse QoS demands, data collection, heterogeneous networks, and multiple resource dimensions; hybrid offline-online learning and parallel computing are identified as needs,.Source: 3) INTELLIGENT AIR INTERFACE • · pp. 59–60

F. BEYOND-COMMUNICATION TECHNOLOGIES FOR 6G

  • Beyond-communication technologies—computation, sensing, and energy—are expected to extend 6G network capabilities and enable novel applications.5G edge computing and enhanced positioning provide a foundation for broader use of these technologies in 6G.Source: F. BEYOND-COMMUNICATION TECHNOLOGIES… · p. 61

1) INTEGRATED COMMUNICATION, COMPUTATION, AND CACHING

  • Integrated communication, computation, and caching (i3C) jointly uses edge capabilities to improve performance and enable services, but remains too immature for large-scale practice.Caching can reduce latency and communication overhead, while edge computing can bring intelligence closer to users; key challenges include resource management, latency, real-time analytics, mobility, security, and privacy,.Source: 1) INTEGRATED COMMUNICATION,… · p. 61

2) INTEGRATED SENSING AND COMMUNICATION •

  • Integrated sensing and communication (ISAC) shares cellular spectrum, infrastructure, hardware, and processing to enable ubiquitous sensing and improve spectrum, energy, and cost efficiency.Mutual assistance can improve beamforming and sensing, but practical ISAC faces full-duplex requirements, limited cellular sensing range, and trade-offs between sensing and communication; communication has design priority,.Source: 2) INTEGRATED SENSING AND… · pp. 62–63

3) WIRELESS ENERGY TRANSFER •

  • Wireless energy transfer (WET) can power devices through near-field coupling or far-field energy beamforming, with far-field efficiency declining over distance.Wireless information and power transfer combines energy and information over the same network through simultaneous transfer, wireless-powered uplinks, or backscatter communication.Source: 3) WIRELESS ENERGY TRANSFER • · p. 65

Past and Present

  • Far-field WET has no large-scale commercial applications, and despite decades of study and some near-field products, the technology remains far from maturity.Near-field examples include wireless chargers, electronic medical implants, and electric vehicles; ITU-R published reports on radio-frequency WET in 2016 and 2021.Source: 3) WIRELESS ENERGY TRANSFER • · pp. 65–66

Opportunities and Challenges

  • WET could enable sustainable 6G IoT and applications including autonomous systems, crowd sensing, and distributed EI, but integration with communication, computation, sensing, and positioning remains challenging,.Key challenges also include efficiency, distance, availability, mobility, and safety.Source: 3) WIRELESS ENERGY TRANSFER • · p. 66

Literature and Future Directions

  • Research should improve RF efficiency and practical transfer distances from meters to tens of meters, while testing higher frequencies, more antennas, and more low-power devices for 6G.RISs may improve efficiency and range, while UAVs may add flexibility and availability; wireless-powered massive lightweight IoT networks are another research direction,,.Source: 3) WIRELESS ENERGY TRANSFER • · p. 66
  • Radiating near-field WET may benefit some higher-frequency, many-antenna scenarios, requiring new research on energy beamforming, SWIPT, and metasurfaces,.Channel estimation, beam design, and waveform design may need rethinking because far-field studies may not apply.Source: 3) WIRELESS ENERGY TRANSFER • · p. 66

G. ENERGY-AWARE TECHNOLOGIES FOR 6G

The section surveys green networks, energy harvesting, and backscatter communications as approaches to energy-aware 6G. It outlines their potential for efficient, sustainable, and lightweight mobile networks alongside practical challenges.

1) GREEN NETWORKS •

  • Green-network design targets energy savings at network, communication, and device levels, with AI/ML expected to support energy-aware management.The section notes that energy-intensive AI/ML itself must also be made lightweight and efficient.Source: 1) GREEN NETWORKS • · pp. 67–68

2) ENERGY HARVESTING •

  • RF energy harvesting can draw on ambient or dedicated radio sources, but remains immature for large-scale cellular use and faces conversion, distance, and cost challenges.Ambient sources avoid dedicated infrastructure but may have limited availability; dedicated sources provide more assured energy levels.Source: 2) ENERGY HARVESTING • · pp. 68–69

3) BACKSCATTER COMMUNICATIONS •

  • Backscatter transmitters encode data by reflecting external RF signals rather than generating their own, supporting low-power, low-complexity communication.Monostatic systems face round-trip pathloss and doubly near-far constraints; ambient systems avoid dedicated sources but can have unstable performance.Source: 3) BACKSCATTER COMMUNICATIONS • · p. 70
  • Backscatter could support energy-constrained IoT, but low data rates, short ranges, uncertain QoS, and security risks remain key challenges.The section says competing goals such as rate, range, robustness, security, energy, and cost require system-dependent trade-offs.Source: 3) BACKSCATTER COMMUNICATIONS • · pp. 69, 71, 72

H. END-DEVICE-ORIENTED COMMUNICATION TECHNOLOGIES FOR 6G

The section surveys D2D and V2X as end-device-oriented technologies for 6G. It describes their potential roles, current capabilities, and challenges in supporting diverse connected environments.

1) D2D COMMUNICATIONS •

  • D2D enables direct communication between nearby devices without base-station involvement, with potential gains in efficiency, latency, throughput, and energy use.Underlay D2D can improve spectral efficiency, while cellular and D2D tiers introduce inter-tier and intra-tier interference-management challenges.Source: 1) D2D COMMUNICATIONS • · pp. 72–73
  • 6G D2D research could extend beyond communication to computation, caching, sensing, and positioning across cellular, industrial, vehicular, and aerial settings.The section identifies THz, UAV, RIS, and AI/ML-aided D2D as potential research directions.Source: 1) D2D COMMUNICATIONS • · pp. 73–74

2) V2X COMMUNICATIONS •

  • V2X includes vehicle-to-vehicle, infrastructure, network, and pedestrian links, supporting goals such as road safety, traffic management, and connected driving.DSRC supports V2V and V2I without network infrastructure; C-V2X also supports V2N connectivity.Source: 2) V2X COMMUNICATIONS • · p. 74
  • 6G V2X research is exploring AI/ML, broad-spectrum and non-terrestrial connectivity, and beyond-communication technologies, while facing design and deployment challenges.The section notes that satellite delays are problematic for time-sensitive V2X, while UAVs can flexibly boost capacity where needed.Source: 2) V2X COMMUNICATIONS • · pp. 75–76

3) CELLULAR-CONNECTED UAV COMMUNICATIONS •

The article surveys cellular-connected UAV communications, private networks, and 6G security as parts of its broader 6G technology guide. It outlines their promise, distinct requirements, and research challenges.

3) CELLULAR-CONNECTED UAV COMMUNICATIONS •

  • UAV-to-everything communications encompass links between UAVs and network nodes, including UAV-to-network, UAV-to-UAV, and UAV-to-device modes.Potential modes also include UAV-to-satellite and UAV-to-vehicle connectivity, spanning ground, cellular, vehicular, and space network layers.Source: 3) CELLULAR-CONNECTED UAV… · p. 76
  • Cellular-connected UAVs could support applications from monitoring and logistics to search and rescue, but require 3D coverage, interference management, and mobility solutions.High-altitude devices, distinctive 3D channels, air-ground interference, dynamic topology, and frequent 3D handovers distinguish UAV communications from ground networks.Source: 3) CELLULAR-CONNECTED UAV… · p. 77

1) PRIVATE NETWORKS •

  • 6G private networks are expected to tailor high-quality wireless services to vertical industries, but must meet customer-specific QoS, coverage, privacy, and security requirements.The article identifies resource management, spectrum use, AI/ML optimization, integrated computation and sensing, fronthaul/backhaul, privacy, and security as challenges.Source: 1) PRIVATE NETWORKS • · pp. 78–79
  • Private networks can be stand-alone or integrated with public networks; stand-alone operation offers customer customization and independence at higher capital and operating costs.5G non-public networks were introduced in Release 16, while private networks remain at an early evolutionary stage.Source: 1) PRIVATE NETWORKS • · p. 79

1) NETWORK SECURITY ARCHITECTURE •

  • A comprehensive 6G security architecture is intended to protect trust and privacy across network layers, through threat identification, solution development, and architecture design.Emerging technologies and applications create new threat vectors, while AI/ML can both defend against attacks and become a target itself.Source: 1) NETWORK SECURITY ARCHITECTURE • · pp. 80–81

X. DEFINING FEATURES FOR 6G

The article characterizes 6G through performance, flexibility, intelligence, availability, sustainability, and security, and summarizes its vision and technologies. It also sketches a post-6G 7G vision centered on a mobile hyperverse.

X. DEFINING FEATURES FOR 6G

  • 6G is defined by 12 features, including extreme performance, flexibility, intelligence and awareness, ubiquitous availability and reliability, sustainability, and security.The article identifies capacity, latency, reliability, density, coverage, and mobility as performance dimensions, with energy efficiency and positioning accuracy also important.Source: §X · pp. 81–82

XI. 6G IN A NUTSHELL

  • The 6G nutshell centers on wireless, artificial intelligence, and Internet of Everything, alongside applications, use cases, requirements, and a range of potential technologies.The listed technology areas include spectrum, antennas, transmission, network architecture, intelligence, beyond-communication, energy-aware, end-device-oriented, service-oriented, and security technologies.Source: §XI · p. 82

XII. 7G VISION: A HIGH-LEVEL SKETCH

  • The speculative 7G vision proposes a mobile hyperverse built on unlimited wireless, deep intelligence, and the Internet of Intelligence.Its envisioned applications span human–deep-intelligence interactions, a deeply intelligent world, and interconnected intelligent ecosystems; realizing it requires mature networks, devices, interfaces, applications, and intelligence.Source: §XII · p. 83

XIII. CONCLUSION

  • The article provides a comprehensive 6G vision, survey, and tutorial, then introduces a high-level 7G vision to place 6G in a wider perspective.Its stated aim is to inspire research and development in academia, industry, and standardization bodies.Source: §XIII · p. 84

XIV. APPENDIX

The appendix provides expansions for abbreviations used in the article, spanning generations, organizations, network technologies, and communication modes.

Appendix

  • The glossary expands generation and organization abbreviations, including 3GPP as 3rd Generation Partnership Project and ITU-R as ITU Radiocommunication Sector.It also defines AI as artificial intelligence and 6G-IA as 6G Smart Networks and Services Industry Association.Source: §XIV · pp. 84–85
  • Network and radio abbreviations include CF-mMIMO for cell-free massive MIMO, CSI for channel state information, and OFDM for orthogonal frequency-division multiplexing.The glossary also expands NFV, NOMA, and NTNs.Source: §XIV · pp. 84–85
  • UAV communication abbreviations include U2D for UAV-to-device, U2N for UAV-to-network, U2S for UAV-to-satellite, U2U for UAV-to-UAV, and U2V for UAV-to-vehicle.The glossary also defines U2X as UAV-to-everything and UAS as unmanned aerial system.Source: §XIV · p. 86

Related papers

Loading 2407.09398v3…