PaperlayerView exact revision on arXiv

Source-linked AI summary

On the Road to 6G: Visions, Requirements, Key Technologies and Testbeds

Cheng-Xiang Wang, Xiaohu You, Xiqi Gao, Xiuming Zhu, Zixin Li, Chuan Zhang, Haiming Wang, Yongming Huang, Yunfei Chen, Harald Haas, John S. Thompson, Erik G. Larsson, Marco Di Renzo, Wen Tong, Peiying Zhu, Xuemin, Shen, H. Vincent Poor, Lajos Hanzo

arXiv:2302.14536v1eess.SP

TL;DR

6G research lacks a settled account of its requirements, scenarios, technologies, and verification platforms while 5G remains limited for emerging applications. This paper surveys and critically appraises the 6G vision, architecture, technologies, testbeds, and open challenges, synthesizing the current research landscape and lessons learned.

Abstract

from arXiv · show

Fifth generation (5G) mobile communication systems have entered the stage of commercial development, providing users with new services and improved user experiences as well as offering a host of novel opportunities to various industries. However, 5G still faces many challenges. To address these challenges, international industrial, academic, and standards organizations have commenced research on sixth generation (6G) wireless communication systems. A series of white papers and survey papers have been published, which aim to define 6G in terms of requirements, application scenarios, key technologies, etc. Although ITU-R has been working on the 6G vision and it is expected to reach a consensus on what 6G will be by mid-2023, the related global discussions are still wide open and the existing literature has identified numerous open issues. This paper first provides a comprehensive portrayal of the 6G vision, technical requirements, and application scenarios, covering the current common understanding of 6G. Then, a critical appraisal of the 6G network architecture and key technologies is presented. Furthermore, existing testbeds and advanced 6G verification platforms are detailed for the first time. In addition, future research directions and open challenges are identified for stimulating the on-going global debate. Finally, lessons learned to date concerning 6G networks are discussed.

I. INTRODUCTION … II. THE GLOBAL 6G VISION

The survey frames 6G as a response to 5G’s capability gaps and synthesizes its emerging global vision, research activity, contributions, and distinguishing technical ambitions. It presents 6G as an integrated, intelligent network extending communication across physical, digital, and societal domains.

I. INTRODUCTION

  • 5G has commercialized at 20 Gbps using sub-6 GHz and mmWave, while advancing antennas, multiple access, and services such as AR, VR, and IoE.Massive MIMO, mmWave, and ultra-dense networking support up to 64 transceiver chains, with commercial products using 128 antennas.Source: §I · pp. 1–2

A. 5G Limitations and Challenges

  • 6G is motivated by 5G gaps requiring global coverage, Tbps rates, sub-millisecond or microsecond latency, 10^8 devices/km^2, centimeter-level positioning, and 100-fold higher network energy efficiency.These challenges also include ultra-reliable and secure connectivity, lower power consumption, and greater intelligence for emerging applications.Source: A. 5G Limitations and Challenges · p. 2

B. Recent Developments

  • Governments, standards bodies, alliances, and industry worldwide have launched 6G programs, with ITU research targeting completion of the 6G vision and technical propositions by 2023.Initiatives span Finland, the United States, China, Japan, South Korea, Germany, Europe, and international standardization organizations.Source: B. Recent Developments · p. 2
  • White papers and surveys define 6G requirements, application scenarios, KPIs, architectures, and enabling technologies, including visions of ubiquitous wireless intelligence and intelligent connection of everything.Organizations including Finland’s 6G Flagship, CCID, Ericsson, IMT-2030, and multiple industry alliances have shaped the emerging discourse.Source: B. Recent Developments · pp. 2–3, §II · pp. 4–5

C. Key Contributions

  • The survey critically appraises the global 6G vision, KPIs, application scenarios, network architecture, enabling technologies, standardization, testbeds, verification platforms, open challenges, and lessons learned.It also proposes a promising network architecture and reviews existing 6G testbeds for wireless channels, components, and verification platforms for the first time.Source: C. Key Contributions · pp. 3–4

II. THE GLOBAL 6G VISION

  • Compared with 5G’s intelligent IoE, 6G targets wider coverage, higher rates, more connections, lower latency, better positioning, integrated communications and sensing, intelligence, security, and substitutability.The vision extends mobile Internet and IoE through deep integration with AI and big data, making 6G a network beyond communication.Source: §II · p. 4

A. Global Coverage … F. Strong Security

The paper presents 6G as an integrated vision spanning ubiquitous space-air-ground-sea connectivity, full-spectrum access, intelligent and multisensory applications, digital-physical integration, and endogenous security. Across these dimensions, 6G combines emerging technologies, standards efforts, and new services to extend communication coverage, capabilities, and trustworthiness.

A. Global Coverage

  • 6G will integrate space-air-ground-sea networks to provide seamless three-dimensional global coverage, including remote access, emergency services, and deep-space environment monitoring.This integration includes satellite, UAV, terrestrial ultra-dense, maritime, underwater, and underground communications.Source: A. Global Coverage · p. 6

B. All Spectra

  • 6G will exploit sub-6 GHz, cmWave, mmWave, THz, and optical bands, combining multiple frequencies to support huge traffic, high data rates, and services where RF communications are limited.THz offers huge bandwidth, while optical wireless bands provide extensive unused spectrum and enable VLC, LiFi, OCC, FSO, and LiDAR.Source: B. All Spectra · pp. 6–7

C. Full Applications

  • 6G will use AI, big data, and integrated communication, computing, sensing, storage, control, localization, and robotics to enable intelligent applications such as smart cities and ISAC.Wireless AI is more mature than ISAC, whose development remains relatively preliminary.Source: C. Full Applications · p. 7

D. All Senses

  • 6G will provide multisensory interconnections by reproducing sight, hearing, touch, taste, and smell through holographic communications, immersive XR, tactile Internet, and related applications.These capabilities target entertainment, skill learning, and medical health, supported by reliable large-bandwidth, low-latency communications.Source: D. All Senses · pp. 7–8

E. All Digital

  • 6G digital twins will precisely map and influence physical reality in real time, enabling predictive decision support and intelligent human-machine-things-environment connections.Representative applications include human body digital twins and digital twin cities, alongside ongoing IEEE and ISO/IEC JTC1 standardization.Source: E. All Digital · p. 8

F. Strong Security

  • 6G must provide strong, endogenous security against traditional and emerging threats through physical- and network-layer protection, quantum communication, blockchain, and AI-enabled autonomous defense.The goal is a credible, manageable, and controllable network, supported by ongoing standardization including 3GPP TR 33.813 and 3GPP TS 33.501.Source: F. Strong Security · p. 8

G. Summary … 3) Service efficiency:

The paper presents 6G as a globally covering, secure network for intelligent, full-sensory, and digital-twin applications, while proposing broader technical requirements and KPIs than 5G. It also identifies target improvements in rate, latency, capacity, coverage, spectrum efficiency, energy efficiency, and cost efficiency.

G. Summary

  • The paper’s 6G vision comprehensively covers current academic and industrial prospects, adding strong security, full-sensory applications, and digital twins as research interests.6G is envisioned as globally covering, enabled by full-spectrum mining and integrated technologies, with endogenous security.Source: G. Summary · p. 8

III. 6G TECHNICAL REQUIREMENTS AND APPLICATION SCENARIOS

  • 6G application scenarios impose distinct communication-system performance requirements, motivating detailed consideration of technical requirements and application scenarios.Source: §III · p. 8

A. 6G Technical Requirements

  • The paper argues that updated 5G KPIs plus new positioning, sensing, security, and intelligence indicators are needed, because existing proposals incompletely and inconsistently characterize 6G.It proposes 17 KPIs, classifies them into four categories, and compares their definitions and reference values with 5G indicators.Source: A. 6G Technical Requirements · pp. 8–9, Table V · p. 10

1) Data rate & Delay:

  • 6G targets a 1 Tbps peak data rate, 10 Gbps user-experienced data rate, 0.1 ms latency, and 1 us delay jitter for demanding applications.THz and optical wireless communications can raise peak rate by more than 50 times over 5G, while deterministic networks support low jitter.Source: 1) Data rate & Delay · p. 9

2) Capacity & Coverage:

  • 6G is expected to provide 10 Gbps/m2 area traffic capacity, 10^8 devices/km2 connection density, three-dimensional global coverage, and coverage exceeding 99%.This extends 5G’s primarily terrestrial coverage and requires balancing user experience, energy efficiency, and coverage cost across areas.Source: 2) Capacity & Coverage · p. 9

3) Service efficiency:

  • 6G spectrum efficiency is expected to triple 5G performance, reaching 90 bps/Hz, while network energy efficiency is projected to increase about 100 times.These indicators evaluate service efficiency at multiple levels and support sustainable next-generation services.Source: 3) Service efficiency · pp. 9–10
  • 6G cost efficiency is projected to reach 500 Gb/$, compared with an estimated 5G value of 10 Gb/$, while user communication spending should remain below 1% of GDP per capita.Cost efficiency captures the trade-off between consumer service quality, provider profitability, and communication expenditure.Source: 3) Service efficiency · p. 10

4) Diversified service evaluation: … 4) Wireless brain-machine interface applications:

The paper characterizes 6G service quality through diversified KPIs and proposes application scenarios extending 5G while adding new combinations and domains. It then illustrates these scenarios through wireless data centers, tactile Internet, digital twins, and wireless brain-machine interfaces.

4) Diversified service evaluation:

  • 6G service evaluation adds mobility, battery life, reliability, positioning, sensing, security capacity, and intelligence level to application-specific KPI requirements.Mobility may exceed 1000 km/h, smart-scenario battery life is expected to reach 20 years, stringent uRLLC permits one erroneous bit per 10 million transmitted bits, outdoor and indoor positioning targets are 1 m and 10 cm, and sensing/imaging resolution targets 1 mm.Source: 4) Diversified service evaluation · p. 11

B. 6G Application Scenarios

  • 6G extends eMBB, mMTC, and uRLLC into feMBB, umMTC, and euRLLC while introducing meMBB, MBRLLC, and muRLLC plus additional scenarios such as digital twins and space-air-ground-sea networks.These scenarios are organized around traditional KPIs together with imaging, positioning, sensing, security capacity, and intelligence level, addressing gaps in earlier classifications.Source: B. 6G Application Scenarios · pp. 11–12

1) Wireless data center:

  • THz’s ultra-large bandwidth could enable ultra-high-rate transmission between cloud servers, realizing cloud-based wireless data centers with wireless storage, transmission, and scheduling.This addresses wired data centers’ complexity, maintenance costs, power consumption, and space requirements.Source: 1) Wireless data center · p. 13

2) Tactile Internet:

  • The tactile Internet digitizes and transmits users’ tactile information so remote robots can synchronize actions, support hazardous work, and potentially perform surgery under ultra-low-latency, ultra-high-reliability communications.Examples include remote medical examinations and mining assistance.Source: 2) Tactile Internet · pp. 13–14

3) Digital twin applications:

  • Digital twins map and influence physical and digital worlds, with AI enabling monitoring, prediction, decision support, and applications spanning cities, human bodies, drug development, and epidemic management.The examples rely on smarter and more advanced 6G combined with cloud, fog, and sensor-layer computing.Source: 3) Digital twin applications · p. 14

4) Wireless brain-machine interface applications:

  • Wireless brain-machine interfaces convert brain signals into device commands for smart-home and medical applications, with demonstrated high-bandwidth, high-precision, low-power control and emerging somatosensory feedback.Future 6G improvements in speed, reliability, latency, perception accuracy, and power consumption are expected to support broader deployment,.Source: 4) Wireless brain-machine interface… · p. 14

5) Holographic communication: … 1) NFV:

The surveyed 6G vision spans immersive applications, scenario-dependent KPI trade-offs, an integrated network architecture, and NFV evolution. Across these areas, the paper emphasizes higher performance, new capabilities, flexible deployment, and unresolved implementation challenges.

5) Holographic communication:

  • 6G holographic communication would transmit high-precision real-time 3D images, requiring very large bandwidth, low latency, and high-resolution information exchange.It could also reconstruct holographic representations of people, objects, and experiences from the past or future.Source: 5) Holographic communication · p. 14

6) Emergency rescue communication:

  • 6G emergency rescue communication would use UAV – and satellite networks, for rapidly deployable, wide-coverage connectivity when terrestrial networks fail or coverage is unavailable.The system must provide large bandwidth and sufficient coverage during the golden 72-hour rescue window while limiting non-terrestrial-network power consumption.Source: 6) Emergency rescue communication · pp. 14–15

7) Immersive XR:

  • 6G immersive XR would digitalize and transmit virtual sight, hearing, touch, smell, and taste, enabling realistic applications including entertainment, telemedicine, and remote industrial control.XR combines VR, AR, and MR, while 6G advances in data rate, bandwidth, latency, reliability, imaging, and sensing are expected to enable fully immersive experiences,.Source: 7) Immersive XR · p. 15

C. System Performance Analysis and Trade-offs between KPIs

  • 6G KPI optimization requires application-specific trade-offs because hardware, propagation, resource, and energy constraints prevent all metrics from reaching their extremes simultaneously.For example, latency, jitter, and reliability outweigh peak rate and connectivity in euRLLC, whereas digital twins demand high data rate, low latency, traffic capacity, connection density, positioning, resolution, and intelligence.Source: C. System Performance Analysis and… · p. 15
  • Research has extensively studied trade-offs among SE, EE, capacity, reliability, latency, security, cost efficiency, and other KPIs, but increasingly complex 6G systems and diverse scenarios still require further analysis.SE enhancement can reduce EE in deployment, making optimal operating points necessary for overall performance.Source: C. System Performance Analysis and… · pp. 15–16

IV. 6G NETWORK ARCHITECTURE

  • 6G network architecture must evolve beyond air-interface improvements to support enhanced and novel applications, including digital twins, integrated communication-computing-sensing, and distributed AI.The proposed architectural directions are 3D multi-network integration; security and trustworthiness; communication, sensing, and computing integration; green, flexible, lightweight design; and native intelligence.Source: §IV · p. 16

1) NFV:

  • NFV decouples software from hardware and uses virtualized servers to provide configurable network functions, reducing cost while improving operational efficiency, flexibility, and scalability,.Its architecture comprises NFVI, VNFs, and NFV-MANO.Source: 1) NFV · p. 16
  • NFV research addresses throughput and latency degradation through acceleration, reinforcement-learning-based scheduling, resource allocation, AI/ML optimization, and security, but deployment still faces complexity, automation, privacy, and virtualization challenges.Further work is needed on cloud-native NFV-MANO, network programming and automation, crowdsourced edge-NFV service quality and privacy, and edge-cloud virtualization.Source: 1) NFV · pp. 16, 18

2) SDN: … 1) 3D multi-network integrated:

The paper surveys 6G network development trends spanning SDN, service-based architecture, network slicing, and integrated three-dimensional heterogeneous networks. These paradigms improve programmability, modularity, customization, resilience, coverage, and intelligent resource management while leaving deployment, security, and coordination challenges.

2) SDN:

  • SDN separates control and forwarding planes through application, control, and data planes, centralizing control to provide programmability, flexibility, lower costs, and efficiency.A software-defined logical network is abstracted over physical infrastructure whose lower-layer devices retain forwarding functions.Source: 2) SDN · p. 18
  • SDN research extends across transport, smart-grid, underwater, satellite, vehicular, edge-computing, blockchain, controller placement, synchronization, and ML-enabled network intelligence.Its logically centralized control and global network view create intelligence opportunities, but security issues and challenges in hybrid forwarding, virtualization, energy efficiency, handover, and freshness-aware multipath control remain.Source: 2) SDN · p. 19

3) SBA:

  • SBA retains cloud computing, virtualization, and micro-services benefits such as on-demand computing, flexible resource management, fine-grained operation, and independent scalability.Its control-plane and user-plane separation parallels SDN’s control-function decoupling, while security concepts and challenges remain under study.Source: 3) SBA · p. 19
  • SBA modularizes 5G core functions into service-connected modules and is evolving toward flexible end-to-end coverage across the RAN, user plane, and integrated air-space-ground networks.A holistic end-to-end SBA is presented as a potential next-generation architecture, although RAN service definitions and module interfaces require further study.Source: 3) SBA · p. 19

4) Network slicing:

  • Network slicing multiplexes independent logical networks over shared infrastructure, but future systems must improve end-to-end integrity, specialization, and intelligence.Key evolution directions are end-to-end slicing, tailored resource allocation, and AI-, digital-twin-, or DRL-enabled management, alongside slice isolation and multi-tenant challenges.Source: 4) Network slicing · pp. 19–20
  • Future network slicing increasingly targets specialized services, including tailored slices and two-level customized RAN slicing that simultaneously satisfies uRLLC and eMBB QoS requirements.AI-assisted architectures, scalable digital twins, and hierarchical DRL are explored for topology, protocol, performance prediction, QoS, spectral efficiency, and resource management.Source: 4) Network slicing · p. 20

1) 3D multi-network integrated:

  • A 6G 3D full-space network will integrate terrestrial, space, aerial, maritime, underwater, and underground networks to improve coverage, resilience, deployment speed, and edge-area operating costs.Space-based networks may provide backhaul, base-station functions, or selected core functions including AMF, UPF, and SMF.Source: 1) 3D multi-network integrated · p. 20
  • Ultra-dense heterogeneous networking is a key 6G trend because integrating multiple technologies and networks can improve overall QoS and reduce costs.Research addresses coverage and throughput self-optimization, adaptive cell selection, movement-aware CoMP handover, and ML/DL methods for ultra-dense networks.Source: 1) 3D multi-network integrated · p. 20

2) Secure and trustworthy: … 5) Natively intelligent:

The paper presents 6G architecture directions spanning endogenous security, integrated communication-computing-sensing, greener and more flexible networking, and native intelligence. These directions combine new architectural models with distributed resources, adaptive operation, and digital-twin support, while remaining subject to substantial research challenges.

2) Secure and trustworthy:

  • 6G security must adopt endogenous, multilateral trust because blurred boundaries and new services create threats that traditional trust models cannot address.The proposed multilateral model combines bridge, consensus, and endorsement trust modes, while security is integrated from the beginning of network design.Source: 2) Secure and trustworthy · pp. 20–21

3) Integration of communications, computing, and sensing:

  • 6G will integrate communication, computing, and sensing across a multi-layer cloud-fog-edge-terminal network, enabling on-demand resource sharing and intelligent services.Nodes are expected to provide data transmission, computing, and sensing, supporting applications such as positioning, mapping, transportation, UAV networking, and environmental detection.Source: 3) Integration of communications,… · pp. 21–22
  • The integrated computing network is progressing from separated communication and computing toward collaboration, but resource demand, computing-power measurement, management, control, and multi-party cooperation remain open challenges.ITU-T Y.2501 defines a computing-power network framework, while current research investigates resource allocation, cloud-edge collaboration, and deployment-oriented operation.Source: 3) Integration of communications,… · pp. 21–22

4) Green, flexible and lightweight:

  • 6G architecture aims to become greener, more flexible, and lighter through cell-free, RAN-Core convergence, and fully decoupled RAN designs.Cell-free massive MIMO jointly serves users through distributed access points and centralized processing, while RAN-Core convergence reduces network complexity and transmission cost; both approaches still require further study.Source: 4) Green, flexible and lightweight · pp. 22–23
  • Fully decoupled RAN separates control and data base stations, including uplink and downlink functions, to improve flexibility, spectrum use, interference control, reliability, mobility, and energy efficiency.Its research remains at an early stage despite proposed uplink base-station selection strategies.Source: 4) Green, flexible and lightweight · p. 23

5) Natively intelligent:

  • Native 6G intelligence consists of AI for Network and Network for AI, enabling self-operation, self-maintenance, self-repair, and more intelligent application services.Cognitive service architectures add real-time perception, AI reasoning, knowledge-graph updates, and intelligent resource scheduling to network functions.Source: 5) Natively intelligent · p. 23
  • DEN2 and self-sustaining networks extend intelligence to collaborative deep-edge nodes and autonomous operation under dynamic or unknown conditions.DEN2 supports native-AI data access, storage, processing, inference, and knowledge distribution, while self-sustaining RAN slicing combines resource self-management, self-optimization, self-learning, and adaptive control.Source: 5) Natively intelligent · p. 24
  • Digital-twin architectures can guide 6G deployment, management, and operation through real-time physical-virtual mapping, but holistic implementations remain preliminary and impose stringent network KPI requirements.Digital twins support closed-loop simulation and optimization, anomaly analysis, slice-resource allocation, and edge-network management, while requiring high data rate, reliability, and low delay.Source: 5) Natively intelligent · pp. 24–25

C. A Novel Promising 6G Network Architecture · V. KEY 6G TECHNOLOGIES

The paper proposes a comprehensive 6G architecture integrating heterogeneous networks, intelligent multi-layer computing, flexible radio and access technologies, native intelligence, endogenous security, and digital twins. It also organizes 16 potential 6G key technologies into four evolution directions and surveys their concepts, applications, developments, and challenges.

C. A Novel Promising 6G Network Architecture

  • The proposed comprehensive 6G architecture unifies heterogeneous future network components, including hierarchical intelligent computing, cell-free networking, and digital-twin-based intelligence.It extends 5G evolution toward an integrated framework covering more potential technologies and components.Source: C. A Novel Promising 6G Network… · p. 25
  • Cloud, fog, and edge computing form a hierarchical intelligent network, while cell-free ultra-massive MIMO, RIS, UDHNs, full-spectrum access, and space-air-ground-sea integration improve coverage, capacity, and efficiency.RAN-core converged edge centers reduce transmission delay, and RAN control and user planes are further decoupled for flexible resource use.Source: C. A Novel Promising 6G Network… · pp. 25–26
  • NFV, SDN, and SBA support end-to-end network slicing, while communication, sensing, computing, and AI integration enables task-centric orchestration and native network intelligence.Network elements gain computing and sensing capabilities, supporting cognitive service architectures, DEN2, and SSNs.Source: C. A Novel Promising 6G Network… · p. 26
  • A multilateral trust model combining blockchain and physical-layer security is intended to provide endogenous security across 6G networks.Data-driven edge intelligent centers integrate communications, sensing, computing, and AI while contributing to native intelligence and security.Source: C. A Novel Promising 6G Network… · p. 26
  • Digital twins will map real 6G networks in real time and use closed-loop simulation to predict optimization performance and guide network deployment, management, and operation.The twin tracks real-network changes and evaluates optimization schemes before guiding the physical network.Source: C. A Novel Promising 6G Network… · p. 26

V. KEY 6G TECHNOLOGIES

  • The paper identifies 16 potential 6G key technologies and groups them into four evolution directions, addressing limitations in prior unclassified or incomplete technology surveys.It also summarizes representative 6G technology surveys in Table VIII.Source: §V · p. 26
  • These directions extend 5G through THz, optical wireless communications, advanced spectrum management, NOMA, ultra-massive MIMO, IBFD, OAM, RIS, and holographic RF, while adding ISAC, AI, blockchain, and semantic communication.The technologies target higher data rates, capacity, efficiency, sensing, intelligence, security, reliability, and improved information extraction; the section discusses their concepts, applications, developments, and challenges.Source: §V · p. 26

A. New Spectrum … 1) New waveforms and modulation:

The paper surveys spectrum expansion and air-interface advances for 6G, spanning THz, optical wireless, and intelligent spectrum management, while motivating scenario-specific waveforms and modulation. These technologies promise higher capacity, efficiency, sensing, and coverage but remain constrained by hardware, channel, propagation, safety, and design challenges.

1) THz:

  • THz deployment still requires suitable high-frequency hardware, accurate channel models, specialized air interfaces and MAC protocols, longer communication distance, and health-and-safety evaluation,,, –,,,.Current CMOS technologies cannot handle frequencies above 300 GHz, while severe propagation loss and power limitations constrain range.Source: 1) THz · p. 27
  • In 2022, a photonics-aided transparent fiber-THz-fiber system exceeded 100 Gbps in real-time transmission, while THz remained some distance from practical applications but was regarded as promising for 6G,,.Source: 1) THz · p. 27

2) OWCs:

  • Optical wireless communications complement RF by offering abundant unlicensed spectrum, ultra-low latency, physical-layer security, zero electromagnetic interference, and deployment simplicity,,,,,.VLC, LiFi, OCC, FSO, and LiDAR support applications including indoor, vehicular, underwater, deep-space, localization, and autonomous-vehicle communications.Source: 2) OWCs · pp. 27, 29
  • Optical bands provide three orders of magnitude more spectrum than RF, but optoelectronic bandwidth, channel anisotropy, orientation, blockage, and safety limit utilization.OLED-based VLC exceeded 1 Gbps, while a silicon photomultiplier achieved BER 10^-3 at 3.45 Gbps; adaptive spatial modulation and NOMA address channel limitations.Source: 2) OWCs · pp. 27, 29

3) Spectrum management:

  • Cognitive radio, symbiotic radio, and dynamic spectrum sharing are proposed to improve 6G spectral and energy efficiency under dynamic, environment-dependent traffic –.Deep Q-learning, ambient backscatter, full-duplex sensing, blockchain, and AI support adaptive access, sharing, security, automation, and decision-making.Source: 3) Spectrum management · p. 29

1) New waveforms and modulation:

  • 6G’s Tbps rates, dense connections, broad coverage, and diverse scenarios require targeted waveform design beyond OFDM’s high spectral efficiency and high peak-to-average ratio.Single-carrier waveforms address high-frequency power challenges, OTFS suits mobility, SEFDM and OVXDM improve spectral efficiency, and ISAC requires joint communication-sensing waveforms, –,,.Source: 1) New waveforms and modulation · pp. 29–30
  • Beyond widely used QAM in LTE and NR, selected and irregular QAM, constellation interpolation, multidimensional modulation, and index modulation target shaping gain, PAPR, and robustness,.Modulation selection strongly affects communication-system effectiveness and reliability.Source: 1) New waveforms and modulation · p. 30

2) New coding: … 1) OAM:

The paper surveys 6G technologies spanning unified coding, NOMA, ultra-massive MIMO, IBFD, and new physical dimensions such as OAM, emphasizing their capabilities and implementation challenges.

2) New coding:

  • Advanced Turbo, LDPC, and polar codes support 6G reliability and efficiency, motivating unified decoding frameworks and circuit-level designs across coding schemes.Short-code scenarios weaken randomness, sparsity, and polarization, motivating OSD, GRAND, PAC, and spatiotemporal 2-D coding alternatives.Source: 2) New coding · p. 30

3) Modern random access:

  • NOMA reuses time, frequency, and/or code resources through successive interference cancellation, improving capacity, spectrum efficiency, latency, and connectivity across power-, code-, and interleave-domain schemes.Its practical deployment still requires low-complexity interference cancellation, stronger security and trust, and a unified 6G framework.Source: 3) Modern random access · pp. 30–31

4) Ultra-massive MIMO:

  • Ultra-massive MIMO uses hundreds or thousands of antennas to improve spectral and energy efficiency, coverage, interference suppression, and three-dimensional positioning across diverse frequencies.Its applications include high-resolution spatial beams, non-terrestrial coverage, and accurate 3D positioning.Source: 4) Ultra-massive MIMO · p. 31
  • Ultra-massive MIMO research addresses near-field and wideband effects, higher-frequency operation, RIS and distributed arrays, AI-enabled management, and integration with space-air-ground-sea networks.Real-time operation and training-data acquisition remain open issues for AI-assisted channel estimation, sounding, beam management, and user detection.Source: 4) Ultra-massive MIMO · p. 31

5) IBFD:

  • IBFD can theoretically double spectrum efficiency by simultaneous same-band transmission and reception, but practical deployment depends on affordable, effective self-interference cancellation, especially at THz and OWC bands.IBFD has been explored for relaying, ambient backscatter, radar-communication, and photonics-assisted cancellation, while suppression difficulty rises with bandwidth.Source: 5) IBFD · pp. 31–32

C. New Physical Dimensions

  • To increase spectral efficiency beyond conventional air-interface dimensions, 6G considers OAM, RIS, and holographic radio as new physical dimensions and transmission carriers.OAM adds an orthogonal angular-momentum mode dimension, allowing simultaneous data transmission over one link without additional frequency bands.Source: C. New Physical Dimensions · p. 32, 1) OAM · p. 32

1) OAM:

  • OAM-based systems improve capacity and spectral efficiency through mode multiplexing and MIMO integration, including throughput gains up to 30.50% and over 200 Gbps transmission demonstrations.OAM is also combined with OFDM, NOMA, and sensing systems for high-rate communication and radar applications.Source: 1) OAM · p. 32
  • OAM deployment remains constrained by beam divergence, misalignment, disrupted orthogonality in reflections and refractions, limited NLoS understanding, and unresolved channel, antenna, component, and signal-processing challenges.These issues currently limit transmission distance and practical commercialization.Source: 1) OAM · pp. 32–33

2) RIS: · 3) Holographic radio: · D. More Capabilities

The section presents RIS and holographic radio as enabling technologies for controllable electromagnetic environments, high-capacity spatial multiplexing, and integrated sensing, while AI, ISAC, blockchain, and semantic communication extend 6G intelligence, sensing, security, and connectivity. It also identifies implementation, modeling, deployment, and learning challenges that remain open.

2) RIS:

  • RIS uses programmable sub-wavelength metamaterials to control electromagnetic amplitudes, phases, polarizations, and frequencies, improving coverage, rates, and efficiency at low cost and energy.RIS can dynamically manipulate incident signals to produce desired reflected or transmitted signals and is relatively easy to deploy.Source: 2) RIS · p. 33
  • RIS supports smart radio environments, massive connectivity, coverage enhancement, relay replacement, green communication, sensing, positioning, reliability, and holographic metasurface technologies.These applications include avoiding coverage holes, boosting cell-edge transmission, and assisting electromagnetic-environment sensing or high-precision positioning.Source: 2) RIS · p. 33
  • RIS research spans channel models, simulators, prototypes, resource allocation, intelligent reflecting transmission, polarization, NOMA, and learning-based beamforming, with reported gains including up to 300% higher energy efficiency than conventional relaying.Studies also report improved coverage, throughput, ergodic rate, and achievable secondary-user rates in simulated and experimental systems.Source: 2) RIS · p. 33
  • RIS deployment remains constrained by hardware, baseband algorithms, network architecture, networking methods, high-frequency cost and energy consumption, and uncertain deployment scale and methods.These issues require further work before broad RIS application.Source: 2) RIS · p. 33

3) Holographic radio:

  • Holographic radio dynamically reconstructs electromagnetic space through holographic interference and continuous-aperture control, enabling spatial multiplexing, extreme spectral efficiency, traffic density, and capacity.Also called holographic MIMO, it links imaging, sensing, and wireless communication to support electromagnetic-space intelligence.Source: 3) Holographic radio · p. 33
  • Holographic radio can exploit interference for energy-efficiency gains and recover transmitting-source spectral holograms, reducing channel-state-information estimation overhead for future industrial, positioning, power, and IoT applications.Its proposed applications include smart factories, high-precision positioning, precise wireless power supply, and massive IoT data transmission.Source: 3) Holographic radio · p. 33
  • Continuous-aperture realization remains difficult, with reconfigurable holographic surfaces and tightly coupled broadband active antennas using UTC-photodetector-coupled arrays as approximate implementation approaches.The UTC-PD approach is described as low-cost and low-power, while holographic-radio theory, channel models, and low-latency reliable processing remain open problems.Source: 3) Holographic radio · p. 33

D. More Capabilities

  • AI, ISAC, blockchain, and semantic communication are positioned to add intelligence, sensing, security, and intelligent connectivity, with AI spanning 6G layers for adaptive optimization and pervasive network intelligence.AI is motivated by growing network data and complexity, but cannot fully replace conventional methods and requires problem-specific learning techniques.Source: D. More Capabilities · p. 34

2) ISAC: … VI. 6G TESTBEDS

The section surveys ISAC, blockchain, semantic communication, energy-neutral devices, and 6G testbeds as enabling technologies and verification foundations for future 6G networks. It also identifies their principal application benefits, technical limitations, and research or standardization challenges.

2) ISAC:

  • ISAC integrates communication and perception to improve the efficiency of hardware, spectrum, time, and energy while supporting 6G’s perceptive-network vision.The systems mutually assist each other and can perceive services, networks, users, terminals, and environmental objects,.Source: 2) ISAC · p. 34
  • ISAC research must address measurement equipment, scenarios, frequency-band selection, channel correlation and modeling, and joint communication-sensing design without interference or collisions.These requirements span hardware architecture, system design, waveform design, and anti-jamming signal processing.Source: 2) ISAC · p. 35

3) Blockchain:

  • Blockchain offers decentralized, transparent, immutable, traceable, and resilient security for distributed 6G devices and infrastructures,,.Applications include secure spectrum sharing, network slicing, MEC resource sharing, massive IoT, IoV, smart grids, and blockchain-empowered federated learning –, –.Source: 3) Blockchain · p. 35
  • Blockchain’s decentralization can cause long latency, inefficient storage, and limited throughput, while fragmented platforms require unified standardization and regulation.Consequently, blockchain may be suitable only for certain application scenarios.Source: 3) Blockchain · p. 35

4) Semantic communication:

  • Semantic communication transmits source meaning rather than requiring error-free bits, potentially improving communication efficiency and reliability and enabling intelligent connections,.It relies on shared semantic knowledge bases and has been studied for image, text, and speech-related end-to-end communication –.Source: 4) Semantic communication · pp. 35–36
  • Semantic communication still lacks reliable semantic elements, effective error-tolerant correction, resource-efficient detection, interoperable models, and comprehensive security mechanisms,.Security challenges include preventing knowledge-graph tampering and protecting private information during storage and recall.Source: 4) Semantic communication · p. 36

5) Energy-neutral devices:

  • Energy-neutral devices harvest ambient or dedicated RF power and communicate through load-modulated backscatter, enabling battery-free applications such as massive sensing and tracking.Their deployment can avoid battery charging or replacement and reduce reliance on toxic battery materials, –.Source: 5) Energy-neutral devices · p. 36
  • Backscatter communication faces a β^2 path-gain link-budget challenge and device-cost constraints, while massive MIMO provides array gain proportional to antenna number and enables practical large-scale deployment, –.Reciprocity-based TDD preserves this scaling with estimated channels, motivating projects such as H2020-REINDEER.Source: 5) Energy-neutral devices · p. 36

A. Testbeds for 6G Channels … B. Testbeds for 6G Key Technologies

The section presents channel testbeds for 6G, including pervasive simulators and channel sounders spanning diverse spectra, scenarios, and applications. It also introduces representative testbeds for designing and verifying 6G key technologies.

A. Testbeds for 6G Channels

  • Channel characterization, measurement, and modeling underpin 6G system design, analysis, evaluation, optimization, and deployment across evolving bands, scenarios, and technologies.Large-scale fading supports link budgeting, planning, optimization, and resource allocation, whereas small-scale fading supports channel estimation, modulation, and coder design; both support capacity, localization, and positioning.Source: A. Testbeds for 6G Channels · pp. 36–37

1) 6G Pervasive Channel Simulator:

  • The pervasive 6G channel modeling theory unifies channel impulse-response expressions and statistical properties across all spectra and scenarios to guide standardized channel-model construction.GBSM is expected to remain the primary standardized approach, with ray tracing and AI/ML models supplementing particular bands and scenarios.Source: 1) 6G Pervasive Channel Simulator · pp. 37–38, Figure 10 · p. 38
  • The 6G pervasive channel model (6GPCM) uses a unified GBSM framework for all spectra, global-coverage scenarios, and full-application channels, while supporting simplified special cases.Its parameter database can combine standardized-model measurements with ray tracing or AI/ML methods, addressing measurement cost and limitations.Source: 1) 6G Pervasive Channel Simulator · p. 38
  • 6GPCM enables analysis of mappings among channel-model parameters, channel characteristics, and system performance, supporting 6G standardization and space-air-ground-sea integrated-network research.The model is described as applicable to sub-6 GHz, mmWave, THz, infrared, and visible-light communications, including LEO satellite, UAV, maritime, ultra-massive MIMO, IIoT, and RIS channels.Source: 1) 6G Pervasive Channel Simulator · p. 38

2) Channel Sounders:

  • Channel sounders actively measure unknown propagation environments and are evaluated using bandwidth, delay range, channel-snapshot repetition rate, and dynamic range.Higher bandwidth improves delay resolution but increases equipment cost or complexity; mobile measurements require a snapshot rate exceeding twice the maximum Doppler frequency.Source: 2) Channel Sounders · pp. 38–39
  • Existing channel sounders cover sub-6 GHz, mmWave, THz, and optical wireless frequencies, with time- and frequency-domain systems supporting SISO, MIMO, and cross-band measurements.Reported systems include 3.5 GHz outdoor measurements, 28 GHz mmWave measurements, 140–220 GHz and 220–330 GHz THz measurements, and infrared or visible-light measurements.Source: 2) Channel Sounders · pp. 39–40
  • Coverage-specific sounders address maritime, UAV, and underground-mine channels, while application-specific systems target massive MIMO, RIS, IIoT, and ISAC scenarios.These systems use scenario-adapted hardware, including lightweight UAV sounders, large antenna arrays for massive MIMO, RIS-integrated platforms, and separate communication and sensing measurements for ISAC.Source: 2) Channel Sounders · p. 40

B. Testbeds for 6G Key Technologies

  • Organizations have proposed representative testbeds for designing and verifying new 6G key technologies, with an overview provided in Table XII.The passage establishes the scope of the subsequent key-technology testbed discussion without specifying individual testbed results.Source: B. Testbeds for 6G Key Technologies · p. 40

1) Testbeds for mmWave: … 5) Testbeds for Cell-free Systems:

The surveyed testbeds validate key 6G technologies across mmWave, THz, RIS, ISAC, and cell-free systems, demonstrating high-throughput links, sensing capabilities, reconfigurable propagation, and scalable distributed architectures. Together, they expose practical implementation challenges while providing platforms for communication, sensing, and networking experimentation.

1) Testbeds for mmWave:

  • mmWave testbeds demonstrated 4.8-Gbps downlink transmission at 28.5 GHz and 1.5 Gbps bidirectional uncompressed high-definition video at 26.5–29.5 GHz.The platforms also addressed mobility and indoor coverage through reconfigurable antenna arrays, automatic beam tracking, and photonics-assisted transmission.Source: 1) Testbeds for mmWave · pp. 40–41
  • Real-time mmWave massive-MIMO testing evaluated performance with static and mobile users across different actual scenarios using a base station with 16 transceiver chains.The LuMaMi28 platform used beamswitchable antenna arrays at the user side.Source: 1) Testbeds for mmWave · pp. 40–41

2) Testbeds for THz:

  • Photonics-aided THz platforms achieved a 352 Gbps single-line rate with 8.6 bps/Hz net spectrum efficiency and supported 103.125 Gbps real-time streaming.The experiments used 325 GHz wired transmission and a 370 GHz transparent fiber-THz-fiber 2×2 MIMO system.Source: 2) Testbeds for THz · pp. 41–42

3) Testbeds for RIS:

  • RIS prototypes demonstrated 26 dB indoor power gain and successful reflected-wave steering without line of sight under different incident angles.Other prototypes achieved 21.7 dBi antenna gain at 2.3 GHz and 19.1 dBi at 28.5 GHz.Source: 3) Testbeds for RIS · p. 42

4) Testbeds for ISAC:

  • ISAC prototypes achieved more than 500 meters detection distance with 100% vehicle-and-person detection accuracy, while outperforming mainstream traffic radar on reported KPIs.The verification used 3GPP 5G millimeter-wave signals with sensing resources not exceeding 15%.Source: 4) Testbeds for ISAC · p. 42
  • THz-ISAC achieved millimeter-level imaging of occluded objects and 240 Gbps line-of-sight transmission over outdoor medium-to-long distances.The prototype operated in the 100–300 GHz band and used virtual aperture assistance.Source: 4) Testbeds for ISAC · p. 42

5) Testbeds for Cell-free Systems:

  • Cell-free testbeds demonstrated 10.185 Gb/s throughput, over 100 b/s/Hz spectrum utilization, 16-stream transmission, and over 200 bps/Hz spectral efficiency.The systems also supported 128 × 128 MIMO demonstrations and 134 µs air-interface calibration between remote radio units and terminals.Source: 5) Testbeds for Cell-free Systems · p. 43

6) Testbeds for OWC: … 2) EM Information Theory:

The paper surveys OWC and comprehensive 6G verification testbeds, then identifies future research challenges spanning fundamental theories, novel channel modeling, and EM information theory. These platforms and theories target 6G’s requirements for high throughput, broad coverage, low latency, intelligence, and integrated communication capabilities.

6) Testbeds for OWC:

  • OWC testbeds span LiFi indoor mobile multiuser networking, point-to-point FSO, and camera-based OCC, leveraging optical devices beyond RF spectrum.OCC offers smartphone-compatible sensing but reaches around 1 Mbps, whereas LiFi and FSO support substantially higher-rate links.Source: 6) Testbeds for OWC · p. 43
  • OWC demonstrations include 40 Gbps multiuser LiFi, 105 Gbps wavelength-division multiplexing, 13.16 Tb/s over 10.45 km, and 100 Gbps over 2.1 km.Additional demonstrations support mobility, handover, multiuser access, LiDAR integration, eye-safe transmission, and deep-space FSO experimentation.Source: 6) Testbeds for OWC · p. 43

C. 6G Comprehensive Verification Platforms

  • The TKµ platform integrates distributed ultra-massive MIMO, RIS, space-air-ground-sea networking, mmWave/THz links, and hierarchical intelligence for 6G validation.It targets Tbps-level peak rate, kbps/Hz-level spectral efficiency, µs-level delay, full coverage, improved positioning, and endogenous security and intelligence.Source: C. 6G Comprehensive Verification… · pp. 43–44
  • A universal 6G prototype platform provides open end-to-end processing, heterogeneous baseband acceleration, visible-light links, communication-AI integration, and cloud-based extensibility.The platform supports joint validation of 6G technologies across research, standardization, and industrialization, with optimization through AVX512 instructions and multithreading.Source: C. 6G Comprehensive Verification… · p. 44
  • A 70 GHz short-range prototype achieved over 10 Gbps throughput, sub-millisecond latency, real-time 4K VR, and less than 560 mW power consumption.Its design combines low-power polar coding, 1-bit ADCs, and adaptive beam sweeping with a high-speed phased array for immersive last-meter connectivity.Source: C. 6G Comprehensive Verification… · p. 44

VII. FUTURE RESEARCH DIRECTIONS AND CHALLENGES FOR 6G

  • Future 6G research must address novel channels, EM information theory, uniform baseband processing, KPI trade-offs, green networking, and development testbeds.These directions are motivated by the 6G vision of global coverage, full applications, strong security, all spectra, all senses, and all digital.Source: §VII · p. 44

1) Novel Channel Research:

  • Novel channel research should move from passive measurement and modeling toward active recognition and control, including pervasive, AI-based predictive, adaptive, and RIS channel models.Traditional methods are costly, incomplete across frequencies and scenarios, computationally demanding, and unable to predict unknown future channel conditions.Source: 1) Novel Channel Research · pp. 44–45

2) EM Information Theory:

  • EM information theory must integrate EM, information, antenna, and propagation theories because 6G continuous-space networks and holographic MIMO exceed individual theories’ scopes,.This integration enables continuous-space channel-state acquisition and capacity calculation while accounting for increasingly inseparable antennas, environments, and propagation channels.Source: 1) Novel Channel Research · p. 45, 2) EM Information Theory · p. 45

3) Unified Baseband Processing: … 6) Immersive Interactive Network:

The paper identifies architectural, methodological, and application challenges on the road to 6G, spanning unified processing, KPI trade-offs, heterogeneous networks, intelligence, security, and immersive services. It emphasizes system-level integration, efficiency, reliability, scalability, and unresolved deployment constraints across diverse scenarios.

3) Unified Baseband Processing:

  • Unified 6G baseband processing should replace isolated module optimization with iterative or joint processing, unified architectures, and automated hardware generation.Iterative receivers exchange soft information among processing modules –, while unified Bayesian processing, VLSI-DSP methods, and AI-assisted design-space exploration,, target efficiency and flexibility.Source: 3) Unified Baseband Processing · p. 46

4) Trade-off between 6G KPIs:

  • 6G research must characterize trade-offs among existing and emerging KPIs across application scenarios, using information theory to guide antenna, coding, delay, reliability, and capacity design.Existing studies emphasize SE, EE, capacity, and delay, while new-KPI and scenario-specific trade-offs remain limited; higher frequencies and ultra-massive MIMO further motivate this analysis.Source: 4) Trade-off between 6G KPIs · p. 46

1) Space-air-ground-sea Integrated Networks:

  • Space-air-ground-sea networks require pervasive channel models and innovations in architecture, mobility, protocols, resource allocation, routing, authentication, energy efficiency, cost, and AI-based optimization.These networks must integrate satellite,, UAV, ocean,, and ground channels while supporting reliable, safe, and seamless heterogeneous connectivity, –.Source: 1) Space-air-ground-sea Integrated… · pp. 46–47

3) AI-enabled Wireless Communication Networks:

  • AI-enabled wireless networks remain early-stage and need shared data, adaptive AI-to-network configuration models, energy-efficient deployment, and stronger privacy and security protections.Advanced AI models can impose substantial energy and environmental costs, while dynamic data collection, transmission, and distribution increase leakage risks, –.Source: 3) AI-enabled Wireless Communication… · p. 47

4) Integrated Communications, Sensing, and Computing Networks:

  • Integrated communication, sensing, and computing networks require capabilities across access and core networks, with new transceiver, processing, topology, MAC, routing, resource-allocation, sensing, and distributed-computing designs.Research spans ISAC waveform and signal processing –, computing offloading and multilayer collaboration,,, plus coordinated sensing and feature fusion.Source: 4) Integrated Communications, Sensing,… · pp. 47–48

5) Endogenous Security Networks:

  • Endogenous security networks can combine blockchain, physical-layer security, mimic defense, and secure multiparty computation, but scalability, resources, practical deployment, and scenario-specific protocols remain unresolved.Mimic defense uses dynamic heterogeneous redundancy and negative feedback, whereas secure multiparty computation requires efficient protocols for varied and malicious environments –.Source: 5) Endogenous Security Networks · p. 48

6) Immersive Interactive Network:

  • Immersive interactive networks target near-real virtual experiences by transmitting visual, auditory, tactile, taste, and smell information, demanding extreme rate, reliability, delay, coordination, intelligence, and security.Key challenges include acquiring and transmitting multimodal data, coordinating concurrent streams, routing, and meeting latency requirements; edge computing, and ML, may improve performance.Source: 6) Immersive Interactive Network · p. 48

7) Digital Twin Network: … APPENDIX

The paper identifies major requirements and unresolved challenges for digital twins, green networks, 6G testbeds, and future 6G development. It concludes that scenario-driven design, coordinated technology planning, basic research, AI integration, sustainability, and international cooperation are central to realizing 6G.

7) Digital Twin Network:

  • Digital twin networks require real-time physical-to-twin mapping, lifecycle management, and closed-loop control, while large-scale topology, channel modeling, equipment compatibility, and data processing remain major challenges.Wireless-channel twins must characterize propagation environments and predict channel changes accurately in real time; large-scale networks also complicate data collection, storage, management, and topology modeling.Source: 7) Digital Twin Network · p. 48

1) Green Networks:

  • Green 6G requires energy-neutral devices, low electromagnetic fields, near-zero-power technologies, and energy-efficient AI models that account for computational complexity, hardware design, and deployment.Relevant technologies include RF energy harvesting, backscattering, low-power computing, and green AI designed to avoid increasing computational cost.Source: 1) Green Networks · pp. 48–49

2) 6G Testbeds:

  • 6G testbeds face hardware, bandwidth, data-storage, synchronization, and channel-scaling constraints across THz, optical, ultra-wideband, ultra-high-speed, and ultra-massive-MIMO scenarios.High-frequency power amplifiers are difficult to manufacture, ADC/DAC sampling limits measurement bandwidth, and switch-based channel expansion increases measurement time while requiring synchronization.Source: 2) 6G Testbeds · p. 49
  • Testbed development is converging toward fusion of technologies, such as ISAC with THz for high-rate communication and accurate sensing, or mmWave with optical access for larger coverage.Selecting suitable technology combinations and constructing corresponding testbeds are identified as future challenges under demanding 6G service requirements.Source: 2) 6G Testbeds · p. 49
  • Comprehensive 6G testbeds must evolve with emerging technologies and provide open architectures with stronger intelligent-cloud processing and storage capabilities.The required platform should support technologies including ISAC, RIS, ultra-massive MIMO, OWC, AI, space-air-ground-sea networks, and edge intelligence, while potentially serving as a public verification platform.Source: 2) 6G Testbeds · p. 49

A. Lessons Learned

  • The lessons learned call for scenario-based design that jointly optimizes competing objectives and activates the most appropriate operating mode for each application.6G must support services beyond communications, including sensing, localization, and over-the-air computing, using tools that identify Pareto-optimal operating points.Source: A. Lessons Learned · p. 49, Figure 13 · p. 50
  • The paper recommends reasonable technology planning, stronger basic research, mutual AI–6G development, green efficiency indicators, and international cooperation toward globally unified standards and roaming.THz, RIS, and cell-free solutions are closer to technical verification, whereas digital twins, semantic communication, and metaverse technologies require further research; channel measurements and modeling support future standardization.Source: A. Lessons Learned · p. 50

B. A Brief Summary

  • The review concludes that 6G will expand global communication services through new scenarios, broad spectral use, pervasive digital intelligence, strong security, testbeds, and continued research on substantial open challenges.The summary covers 6G vision, KPIs, application scenarios, trade-offs, architectures, technologies, testbeds, green networks, fundamental research, and associated research directions.Source: §VIII · p. 49, B. A Brief Summary · pp. 50–51

Related papers

Loading 2302.14536v1…