Content uploaded by Lauri Lovén
Author content
All content in this area was uploaded by Lauri Lovén on Sep 24, 2019
Content may be subject to copyright.
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 1
KEY DRIVERS
AND RESEARCH
CHALLENGES
FOR 6G
UBIQUITOUS
WIRELESS
INTELLIGENCE
6G Research Visions 1
September 2019
2 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
TABLE OF CONTENTS
3 EXECUTIVE SUMMARY
4 INTRODUCTION
7 SOCIETAL AND BUSINESS DRIVERS FOR 6G
12 6G USE CASES AND NEW DEVICE FORMS
14 6G SPECTRUM AND KPI TARGETS
18 RADIO HARDWARE PROGRESS AND CHALLENGES
22 PHYSICAL LAYER AND WIRELESS SYSTEM
26 6G NETWORKING
29 NEW SERVICE ENABLERS
33 CONTRIBUTORS
6G Research Visions 1
Key Drivers and Research Challenges for 6G Ubiquitous Wireless Intelligence
Mai Latva-aho, Kari Leppänen (eds.)
6G Flagship, University of Oulu, Finland
September 2019
ISBN 978-952-62-2353-7 (print)
ISSN 2669-9621 (print)
ISBN 978-952-62-2354-4 (online)
ISSN 2669-963X (online)
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 3
EXECUTIVE SUMMARY
Our future society will be increasingly digitised, hyper-connected and globally data driven. Many widely
anticipated future services, including eHealth and autonomous vehicles, will be critically dependent on
instant, virtually unlimited wireless connectivity. Mobile communication technologies are expected to
progress far beyond anything seen so far in wireless-enabled applications, making everyday lives smoother
and safer and dramatically improving the efficiency of businesses.
As fih generation (5G) research is maturing towards a global standard, the research community must
focus on the development of beyond-5G solutions and the 2030 era, i.e. 6G. It is not clear yet what 6G will
entail. It will include relevant technologies considered too immature for 5G or which are outside the defined
scope of 5G. More specifically, the way in which data is collected, processed, transmied and consumed
within the wireless network will be a key driver for 6G.
e first 6G Wireless Summit in March 2019 launched the process of identifying the key drivers, research
requirements, challenges and essential research questions related to 6G. is white paper is the first version
for the annually revised series of 6G research visions and can be phrased in one vision statement from the
first 6G Wireless Summit: Ubiquitous wireless intelligence.
It is envisioned that we will need new KPI drivers besides the current 5G technical KPIs. Societal megatrends,
United Nations (UN) sustainability goals, lowering carbon dioxide emissions, emerging new technical
enablers as well as ever increasing productivity demands are critical drivers towards 2030 solutions.
Totally new services such as telepresence and mixed reality will be made possible by high resolution imaging
and sensing, accurate positioning, wearable displays, mobile robots and drones, specialized processors,
and next-generation wireless networks. Current smart phones are likely to be replaced by pervasive XR
experiences with lightweight glasses delivering unprecedented resolution, frame rates, and dynamic range.
6G research should look at the problem of transmiing up to 1 Tbps per user. is is possible through
the efficient utilization of the spectrum in the THz regime. Extended spectrum towards THz will enable
merging communications and new applications such as 3D imaging and sensing. However, new paradigms
for transceiver architecture and computing will be needed to achieve these – there are opportunities for
semiconductors, optics and new materials in THz applications to mention a few.
Artificial intelligence and machine learning will play a major role both in link and system-level solutions of 6G
wireless networks. New access methods will be needed for truly massive machine-type communications.
Modulation and duplexing schemes beyond Quadrature Amplitude Modulation (QAM) and Orthogonal
Frequency Division Multiplexing (OFDM) must be developed and possibly it is time to start looking at
analogue types of modulation at THz frequencies.
Security at all levels of future systems will be much more critical in the future and 6G needs a network with
embedded trust. e strongest security protection may be achieved in the physical layer. During the 6G era
it will be possible to create data markets, and thus, privacy protection is one key enabler for future services
and applications.
6G is not only about moving data around – it will become a framework of services, including communication
services where all user-specific computation and intelligence may move to the edge cloud. e integration
of sensing, imaging and highly accurate positioning capabilities with mobility will open a myriad of new
applications in 6G.
4 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
INTRODUCTION
e arrival of the 5G mobile communications technology is already showing signs of becoming a major
factor in driving productivity and is expected to be the key enabler for long-envisaged, highly integrated and
autonomous applications in many sectors. is new wave of technology will accelerate the digitalisation of
economies and society. Historically, a new mobile “generation” appears approximately every ten years, with
6G expected to emerge around 2030. e first release of 5G New Radio (NR) – 3GPP Release 15 – was
ready in 2018, and global commercialization of 5G is currently taking off. 5G performance and use cases will
continue to evolve in the coming releases. 6G will take onboard new technologies and satisfy communication
demands going beyond the 5G evolution. Now is the perfect time to identify future communication needs,
performance requirements, system and radio challenges, and major technical options for 6G to establish
the research goals towards the 2030s.
e first 6G Wireless Summit1 was organized in Levi, Finland, in March 2019 with almost 300 participants from
29 countries, including major infrastructure manufacturers, operators, regulators as well as academia. e
event was organised by the Finnish 6G Flagship Programme2. e 6G vision statement captures the essence
of many of the key messages from the event: Ubiquitous Wireless Intelligence; Ubiquitous – services follow
users everywhere seamlessly; Wireless – wireless connectivity is part of critical infrastructure; Intelligence –
context-aware smart services and applications for human and non-human users.
Following the summit, a workshop was organized with 70 selected participants to commence the draing
of the first 6G white paper. Each year, the white paper will be updated following the annual 6G Wireless
Summit. e goal for this first edition was to identify the key drivers, research requirements, challenges
and essential research questions related to 6G. e format of the white paper is deliberately short avoiding
lengthy background and justifications; it is targeted primarily at technical experts working in the field. At
the highest level, the workshop identified major drivers for 6G (Figure 1): sustainability, society, productivity
and technology.
Is it naïve to say “From 5G Engineering to 6G Humanity”
or is it imperative?
In 2016, the UN released 17 Sustainable Development Goals3 (SDGs) for the 2030 Agenda. ese goals
were developed against a backdrop of a growing and ageing global population, increasing urbanization
and a world in which the climate is changing. If adhered to, the UN SDGs are expected to drive policy and
influence government spending in many economies, creating global demand. It is estimated that the world’s
population in 2019 is 7.6 billion people, and that this will grow to 8.5 billion by 2030, 9.7 billion by 2050
and 11.2 billion by the end of the current century. As of 2018, 55% of the world’s population lives in urban
areas, a proportion that is expected to increase to 68% by 20504. By 2030, the world is projected to have
43 megacities with more than 10 million inhabitants, most of them in developing regions. However, some
of the fastest-growing urban agglomerations are cities with fewer than 1 million inhabitants, many of them
located in Asia and Africa.
1www.6gsummit.com 2www.6gflagship.com
See hps://www.un.org/sustainabledevelopment/sustainable-development-goals/.
4Source UN Department of Economic Affairs 2018 Revision of World Urbanization Prospects.
Available online hps://www.un.org/development/desa/en/news/population/2018-revision-of-world-urbanization-prospects.html.
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 5
Figure 1. From 5G Engineering to 6G Humanity – Breaking down the four areas driving 6G research.
Urbanization calls for super-efficient ICT services throughout society, which will become more and more
automated in all sectors to significantly increase productivity, reduce carbon dioxide emissions and
generate cost savings for public expenditure. e future services must be available 24-7, ubiquitously. e
services developed for the future urban areas need to be transformed for the needs of remote, oen rural
and very poor areas in order to tackle the UN SDGs. At the same, as societies become heavily dependent on
ICT services, they become extremely vulnerable to various types of security threats and aacks. e global
threats5 can no longer be ignored when developing future 6G technologies. Further to the societal and SDG
drivers, we have included some examples of technology trends and drivers for increased productivity. 5G
is envisioned to solve our communication challenges set for 2020s and beyond. However, the first 5G NR
release only covers a subset of 5G targets and envisioned use cases.
New requirements and technologies are continuously emerging. Some are expected to enter future
releases of 5G whereas far more rich requirements and technologies will need to wait for a clean slate of
6G specifications. Some of the emerging and promising directions in technology, described further in later
chapters, are listed in Figures 2 and 3.
6G
HUMANITY
RAN Agnostic/Automatically Orchestrated
Transceivers • Non-device Centric Communications
HBI • Extreme URLLC • Below CM Positioning •
Consent and Privacy Preserving Data Sharing •
Support for Ambient/Novel Sensing • Small Data AI
(Distributed Learning) • Distributed Trust •
Cyber-physical Security • Terahertz Technologies •
4D-Imaging and Image Projection and XR • Haptic
Remote Telepresence • Full Spectrum Photonic
Signal Processing • Proactive Decision
Making/Informations Oering • Per vasive User
Identification and Authentication • Net Neutrality •
Zero-energy Communications • AI Inspired Air
Interfaces • Grant Free Access (IoT)
Education Innovations • Societal Services •
Health and Wellbeing Services • Urbanisation vs.
Remote • Infrastructure • Work Life Change •
Data Security and Privacy • Automation •
Personalisation
Quality Education • Clean Water and Sanitation •
Gender Equality • Life Below Water • Life on Land •
No Poverty • Good Health and Well-being • Climate
Action • Sustainable Cities and Communities •
Peace, Justice, and Strong Institutions • Clean
Water and Sanitation • Zero Hunger • Industry,
Innovation and Infrastructure • Aordable and
Clean Energy • Reduced Inequalities •
Partnerships for the Goals • Responsible
Consumption and Production • Decent Work and
Economic Growth
Health • Manufacturing • Finance Technologies •
Society 5.0 • Transport • Global Aordable
Coverage • Education • Agriculture • Energy •
FinTech
SUSTAINABILITY GoalsTECHNOLOGY Enablers
PRODUCTIVITY in Vertical IndustriesSOCIETAL Challenges
5Source World Economic Forum: e Global Risks Report 2019. Available online hp://www3.weforum.org/docs/WEF_Global_
Risks_Report_2019.pdf.
6 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
Figure 2. New wireless hardware and physical layer technologies.
Figure 3. Possible technologies for user interface and service enablers.
In the remainder of this white paper, key areas for investigation are identified to make the 2030 vision for
Ubiquitous Wireless Intelligence a reality. e goal is also to identify essential research questions within the areas
of interest. It is acknowledged that this does not form a comprehensive list, rather a starting point reflecting the
discussions at the first 6G Wireless Summit as well as views from the 6G Flagship programme. Future editions of
this white paper will complement the missing areas not discussed at the summit.
HIGH IMPACTLOW IMPACT
LOW UNCERTAINTY HIGH UNCERTAINTY
Hardware
PHY & Wireless System
LOSSLESS
MATERIALS
AT THZ
VISIBLE
LIGHT
COMMU-
NICATION
CMOS IN THZ
PHY
SECURITY
GRAPHENE
“ZERO
ENERGY”
RADIOS
ANALOG
MODULATION
ML ON PHY
LAYER
SIGNAL
SHAPING
INP
AT THZ
BACK-
SCATTER
COMMUNI-
CATIONS
MASSIVE
ANTENNA
ARRAYS SIGE BICMOS
IN THZ
THZ
COMMU-
NICA-
TIONS
HIGH IMPACTLOW IMPACT
LOW UNCERTAINTY HIGH UNCERTAINTY
UI Technologies
Service Enablers
NEURAL
INTERFACES
ALL-DAY
WEARABLE
DISPLAYS
MOBILE
MIXED
REALITY
FOVEATED
RENDERING
MOBILE VR
USER APPS
RUN IN EDGE
LOCAL DATA
MARKET
EDGE
FUSION OF
LOCAL AND
USER
CONTEXT
RADIO
SENSING &
IMAGING
CM
POSITIONING
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 7
SOCIETAL AND BUSINESS DRIVERS FOR 6G
5G was primarily developed to address the anticipated capacity growth demand from consumers, as well
as the productivity demands from industry, and to enable the increasing importance of Internet of ings
(IoT). e technical success of 5G has relied on new developments in many areas and will deliver a much
wider range of data rates to a much broader variety of devices and users. 6G will require a substantially
more holistic approach to identify future communication needs, embracing a much wider community to
shape the requirements of 6G. is includes identifying the trends, demands and challenges facing future
societies, as well as the global forces shaping our future world to avoid merely commercially driven system
definitions. Even though 5G development was shaped by demands from a range of vertical industry sectors,
the emphasis has remained on deployments driven by mobile network operators (MNOs). 6G will introduce
super-efficient short-range connectivity solutions that are likely to be driven by new players in the market
resulting in new ecosystems outside traditional MNOs. Having a more inclusive view outside of MNOs will
help shape the needs of 6G.
Drivers from society, including the UN sustainability goals, will shape 6G.
Societal and business drivers will increasingly shape 6G development, including political, economic,
social, technological, legal and environmental (PESTLE) drivers as highlighted in Figure 4. To ensure that
the benefits of smart city services and urbanization are fully shared and inclusive, policies to manage
urban growth need to ensure access to infrastructure and social services for all, focusing on the needs
of the urban poor and other vulnerable groups for housing, education, health care, meaningful work and
a safe environment. e rise of always-connected omni-present systems, gadgets and sensors serving
digital automation of critical processes will set high requirements for trustworthiness and resilience. e
ubiquitous connectivity and contextual awareness of 6G networks is expected to promote ICT accessibility
and use for the social and economic development of people with specific needs, including indigenous
people and people living in rural areas. Future 6G architectures will foster digital inclusion and accessibility
also unlocking rural economic value and opportunities.
High energy efficiency to reduce the overall network energy consumption is a critical requirement for 6G.
e choice, use, reuse and recycling of materials throughout product lifecycles will enable the total cost
of ownership to be reduced, facilitate the extension of network connectivity to remote areas, and provide
network access in a sustainable and more resource-efficient way. Extensive research has been conducted
into possible health effects of exposure to many parts of the frequency spectrum including mobile
phones and base stations. All reviews conducted so far have indicated that exposures below the limits
recommended in the ICNIRP (1998) EMF guidelines6, covering the full frequency range from 0-300 GHz,
do not produce any known adverse health effects (UN WHO7). e introduction of novel 6G technologies
will initiate the need to review the status of the science and identify gaps in knowledge needing further
research to make beer health risk assessments.
6ICNIRP guidelines for limiting exposure to time-varying electric, magnetic and electromagnetic fields (up to 300 GHz) published in:
Health Physics 74 (4):494-522; 1998. hps://www.icnirp.org/cms/upload/publications/ICNIRPemfgdl.pdf.
7WHO - World Health Organization. Extremely low frequency fields. Environmental Health Criteria, Vol. 238. Geneva, World Health
Organization, 2007. hps://www.who.int/peh-emf/en/
8 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
Figure 4. 6G PESTLE (Political, Economic, Social, Technological, Legal and Environmental) analysis results highlight inclusion, sustainability and transparency.
We are moving towards a data sharing / data market economy
where issues with data ownership and contractual policies require
special aention.
Access to data and data ownership are increasingly major factors in value creation, and limiting such access
is a means of control. Creating a system that transforms how data is collected, prioritized, and shared
can create strong drivers for future value, but may also lead to serious privacy and ethical concerns over
the location and use of data. Furthermore, how the data itself can be used becomes a key question. e
contractual rights and obligations of the different members of a communications ecosystem may describe
how the information and data may be used. e challenge, however, will be the mapping of these rights and
obligations to the data collected and used by highly adaptive autonomous systems, or smart devices, to
create the services of the future.
e transition to ever higher frequencies with smaller radio ranges
and the increasing role of indoor networks will boost network
sharing in cities and indoor spaces, and – especially – drive the
“local operator” paradigm.
For the foreseeable future, operation in the lower frequency bands (below 4 GHz), currently used for
mobile communication networks, are expected to remain stable and dominated by MNOs8 due to long-
term spectrum licenses. With 6G however, new bands that target super-efficient short-range networks for
both indoor environments and outdoor city spaces will become common (Figure 5). ese local networks
will target verticals with specialized demand and will be deployed by different stakeholders opening the
HIGH MED LOW IMPACT LOW MED HIGH
T
P
E
S
L
E
Privacy
AI/ML Rights
Inclusion
Sharing Economy
Rural Economics
Smart City
Resilience
Awareness
Data Ownership
Contracting
Resource Eciency
EMF
AI/ML
Security
PESTLE – Inclusion, Sustainability & Transparency
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 9
market to new players, new investments and new ecosystems9. Building several overlapping ultra-dense
networks will rapidly become infeasible and will lead to different stakeholders deploying a single network
within a facility to serve multiple user groups and services. Via sowarization and the virtualization of
network functions and opening of interfaces, sharing economy concepts will be utilized not only for the high
platform business layers but also widely in network connectivity and data context layers. Challenges related
to prioritising of traffic continue, as in the network neutrality regulation debate. Changes in the ownership
of spectrum access rights, networks, network resources, facilities and customers will result in several
combinations as different facilities will have varying requirements and infrastructures. New incentives will
arise partly through regulation. e global harmonization of the spectrum will still be a challenge to be
solved with the joint effort of all stakeholders.
6G will penetrate deeper into society and lives of people than anything we have seen so far. It will be very
complex and besides communication deals with data collection, processing and ubiquitous intelligence. To
avoid excessive operating costs, 6G soware will be run on cloud technology utilising a very high level of
automation. is will require advances in regulation.
Figure 5. Drivers toward local operator paradigm.
LOCAL
OPERATORS
MULTI-SERVICE MULTI-TENANT NETWORK
FLEXIBLE AND AUTOMATED
NETWORK INFRASTUCTURE
TECHNOLOGY
ALLOWS DIFFERENTIATED
SERVICE QUALITY
HIGHER CARRIER FREQUENCIES
INCREASED
DENSIFICATION COST
COST
SINGLE SHARED
LOCAL NETWORK
LOCAL VERTICAL SPECIFIC NEEDS
DOMAIN SPECIFIC
KNOWLEDGE NEEDED
DOMAIN SPECIFIC
OPERATOR
COMPETENCE
INCREASING ROLE OF DATA
CONCERNS ON DATA
OWNERSHIP AND USAGE
BUSINESS MODELS
DRIVEN BY
LOCAL DATA
DATA
8A. Weber, D. Scuka, Operators at crossroads: Market protection or innovation?, Telecommunications Policy, Volume 40, Issue 4,
2016, Pages 368-377.
9P. Ahokangas, M. Matinmikko-Blue, S. Yrjölä, V. Seppänen, H. Hämmäinen, R. Jurva, M. Latva-aho, "Business Models for Local 5G
Micro Operators," IEEE Transactions on Cognitive Communications and Networking, 2019.
10 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
Stakeholder roles in 6G will change compared to the current
mobile business ecosystem and new roles will emerge.
As a sharing economy, future 6G ecosystems will change the existing roles and introduce new stakeholder
roles resulting in a complex ecosystem as outlined in Figure 6. While the MNO market dominance is expected
to continue in 5G, future 6G connectivity solutions will be driven by new players in the market. Stakeholders
representing the different demands and needs originating from human and machine users, and specific to
the public sector or enterprises in different verticals, are shown in the middle quadrangle in Figure 6. e
resources and assets needed to meet the range of needs will be provided by different stakeholder roles
providing physical infrastructure (facilities, sites), equipment (devices, networks), and data (content, context),
under the regulatory framework set by the policy makers as depicted in the outer quadrangle. Demands and
resources are brought together through the matching/sharing stakeholder roles10 including different kinds of
operators (local or vertical-specific operators, fixed operators, mobile network operators, satellite operators),
resource brokers, and various service/application providers, such as trust/security providers.
Figure 6. Stakeholders in a future 6G ecosystem.
MACHINES
ENTERPRISE
PUBLIC SECTOR
END USERS
MOBILE
NETWORK
OPERATOR
FIXED
OPERATOR
VERTICAL
SPECIFIC SERVICE
PROVIDER
SATELLITE
OPERATOR
ROAMING
SERVICE
PROVIDER
APPLICATION
PROVIDER
DIGITAL
TWIN
PROVIDER
MANAGEMENT
SERVICE PROVIDER
DATA
BROKER
NETWORK
RESOURCE
BROKER
BROKING/
BRIDGING
PROVIDER
TRUST
PROVIDER
SECURITY AS
A SERVICE
BUILDING
CONSTRUCTOR
SITE
SUPPLIER
FACILITY
OWNER
DATA
CENTER
EDGE
CLOUD
CONTEXT
OWNER
CONTEXT
PROVIDER
CONTENT
PROVIDER
DATA OWNER GOVERNMENT
PUBLIC
SECTOR
NATIONAL
REGULATOR
COMPLEMENTARY
TECHNOLOGY
PROVIDER
NETWORK/
CLOUD INFRA
VENDOR
DEVICE
SUPPLIER
MOBILE VIRTUAL
NETWORK
OPERATOR
RESOURCES/ASSETS
DEMANDS/NEEDS
MATCHING/SHARING
10R. Amit, X. Han, Value Creation rough Novel Resource Configurations in a Digitally Enabled World. Strategic Entrepreneurship
Journal, 11, 2017, 228-242.
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 11
Overall, the stakeholder roles in 6G are expected to change compared to the current mobile business
ecosystem and totally new roles will emerge. It is expected that especially the drivers listed in Figure 5 will
fundamentally change the ecosystem and open new opportunities for different kinds of stakeholders in 6G.
e matching and sharing of resources to meet the demands will take place through new activities to ensure
inclusion, sustainability and transparency. Ultimately, the emergence and shape of new 6G ecosystems will
be dependent on regulations which promote or hinder these developments.
Research questions
Q1: What are the key societal requirements for 6G?
Q2: How can ecosystem stakeholders and roles, and their ecosystem configurations be categorized in 6G?
Q3: How, why and what platform-based ecosystem business models could emerge in the 6G sharing economy?
Q4: How could artificial intelligence (AI) and machine learning (ML) transform the platform-based
ecosystems, business models and services in future 6G systems?
Q5: What is the minimum viable regulation needed for 6G to respond to societal requirements?
Q6: How can novel mechanisms and business models be developed which support inclusion,
and access in remote areas?
12 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
6G USE CASES AND NEW DEVICE FORMS
While smartphones have become an indispensable part of our lives, rapid advances in new display
technologies, sensing and imaging devices, and low-power specialized processors are ushering in a new
era in which our devices will become seamlessly integrated with our senses and motoric control. Virtual
(VR)11, 12 augmented (AR)13, and mixed reality (MR) technologies are merging into XR, which encompasses
wearable displays and interaction mechanisms that create and maintain perceptual illusions. Users will
accept an alternative version of reality that enhances their ability to consume media, search the Internet,
explore real and virtual worlds, collaborate on work projects, connect with family and friends, and engage
in restorative activities.
Smart phones are likely to be replaced by pervasive
XR experiences through lightweight glasses delivering
unprecedented resolution, frame rates, and dynamic range.
XR experiences are likely to be delivered by lightweight glasses that project images onto the eyes at an
unprecedented resolution, frame rate, and dynamic range. Furthermore, feedback will be provided to other
senses via earphones and haptic interfaces. e necessary supporting technologies include: 1) imaging
devices such as light field, panoramic, depth-sensing, and high-speed cameras; 2) biosensors for monitoring
health conditions such as the heart rate, blood pressure, and neural activity; 3) specialized processors for
computer graphics, computer vision, sensor fusion, machine learning, and AI, either in the device or in the
surrounding network infrastructure; 4) wireless technologies including positioning and sensing. Sensing and
imaging devices can capture our entire life experiences as well as detailed physical environments, whereas
virtual worlds continue to increase in fidelity. ese advances in combination with the requirement to distribute
computation (because the computation demand exceeds small devices such as glasses) highlight demand
for performance from wireless networks that is not yet available.
Telepresence will be made possible by high resolution imaging
and sensing, wearable displays, mobile robots and drones,
specialized processors, and next-generation wireless networks.
For centuries people have looked for ways to feel closer over great distances. From the postal service to the
telegraph to the telephone to video chaing, our expectations for remote communication and interaction
continue to evolve. Telepresence, as a surrogate for actual travel, is finally becoming a reality with the
unprecedented speed of development in the supporting technologies: high-resolution imaging and sensing,
wearable displays, mobile robots and drones, specialized processors, and next-generation wireless networks.
A sense of presence may be achieved through real-time capture, transmission, and rendering of a 3D
holographic representation of each participant in a meeting, or by combinations of graphical representations,
such as avatars, and movement data that is captured by sensors. Perceptual illusions are created by XR
devices that convince a geographically distributed group of people that they are in the same location, which
could be a real or virtual environment.
11J. Bailenson, Experience on Demand, What Virtual Reality Is, How It Works, and What It Can Do, W. W. Norton and Company, 2018.
12S. M. LaValle, Virtual Reality, Cambridge University Press, 2019.
13D. Schmalsteig, T. Höllerer, Augmented Reality: Principles and Practice, Mendeley Ltd., 2015.
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 13
With autonomous systems and robots, people can even effect changes in the remote world. People may
communicate through familiar mechanisms such as speech and body language, or through specialized
devices, analogous to a PC mouse, optimized for comfort, precision, and efficiency in 3D worlds14. Exemplary
applications include education over distances, collaborative design, telemedicine, telecommuting, advanced
3D simulation and training, and defence.
Autonomous vehicles for ecologically sustainable transport and
logistics are made possible by advances in wireless networks and
in distributed AI and sensing.
Even with the advent of telepresence, the movement of people and goods remains a critical challenge as both
population and globalization increase. e world in 2030 and beyond envisions many millions of networked
autonomous vehicles operating with different degrees of coordination to make transport and logistics as efficient
as possible. ese vehicles may include autonomous cars that move people between their homes and workplaces
or schools, and autonomous trucks or drones that deliver goods. By 2030, online consumer shopping is expected
to dominate in developed countries, causing a need to deliver millions of packages from warehouses to individual
homes. Efficiency is important not only for improved global productivity but also for achieving sustainability
targets by reducing fossil fuel consumption. Even more pressing than efficiency is safety: with the rise of the use
of autonomous vehicles there should be no rise in harm to humans. In fact, the goal should be to reduce today’s
current global death and injury rates from transport and logistics networks. Advances in sensors, sensor fusion,
and control systems continue to improve safety, but this comes at the expense of stronger network requirements.
Each vehicle in a future network will be equipped with many sensors, including cameras, laser scanners, possibly
THz arrays for 3D imaging, odometry, and inertial measurement units. Algorithms must quickly fuse data from
multiple sources and make rapid decisions about how to control the vehicle while considering a locally generated
map of its immediate environment, its place in that environment, information on other vehicles, people, animals,
structures or hazards that might lead to collision or injury. Interfaces must also be developed that alert passengers
or supervisors to potential risks so that appropriate actions can be taken to avoid accidents. For the network of
vehicles to function efficiently and safely, wireless networks must deliver ultra-high reliability, in addition to low
latencies and a high bandwidth. A more comprehensive set of use cases are presented in the 6G Flagship Vision
Video15.
Research questions
Q1: What are the functional, performance and ergonomic requirements for next generation XR systems?
Q2: How to distribute computation and data between different components in future XR systems?
Q3: How can human perception-based quality of experience (QoE) criteria be defined and measured for
next generation XR devices?
Q4: What novel opportunities can next generation networks and devices offer to interaction between
people?
Q5: What are the considerations for communication reliability and traffic safety for autonomous vehicles?
14J. J. LaViola Jr., E. Kruijff, R. P. McMahan, D. Bowman, I. P. Poupyrev, 3D User Interfaces: eory and Practice, 2nd Ed., Addison-Wesley, 2017.
15hps://www.youtube.com/watch?v=T6ubRoZCeVw&list=PLFhJqSbEBkRSbtfHKX5ju0tsGt3bId-nM&index=2&t=0s
14 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
6G SPECTRUM AND KPI TARGETS
Many of the Key Performance Indicators (KPIs) used for developing current and emerging 5G technologies
are valid also for 6G. However, the KPIs must be critically reviewed and new KPIs must be seriously considered.
For technology driven KPIs, some leading vendors have released their initial dras for the Beyond 5G and
6G requirements as depicted in Figure 7.
Figure 7. Generic 6G targets presented by academia and industry in different fora.
e beyond 5G (B5G) and 6G targets in most of the technology domains once again point to an increase in
the respective capability by a factor of 10-100, in line with the previous mobile cellular generation upgrades.
Some of the potential 6G KPIs including the previously discussed technology KPIs are listed in Figure 8.
ere are several KPI classes that are currently difficult to define and will be improved in future releases of
the white paper.
100 Gbps - 1 Tbps
Peak Data Rates
10 cm (indoor),
1 m (outdoor)
Positioning
10x
More Energy
Ecient
Max. 1 Out of
Million Outage
Extreme Ultra
Reliability
10 000x
Trac Increase
100 Devices
per m3
Density
20 Years
Battery Life Time
0.1 ms
Radio Latency
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 15
Figure 8. Initial 6G KPIs.
An example of assessing these new metrics is related to UN SDG KPIs. Some activities have already been
performed16, 17, but further studies are encouraged to define the KPIs and to set target values. Digitalization
accelerates progress towards achieving the SDGs – but business-as-usual will not achieve the SDGs on
time. It is estimated that carefully designed digital solutions, including next generation wireless, targeting
SDG KPIs will offer the required scale and speed for breakthroughs.
6G research should include the challenge of transmiing
up to 1 Tbps per user.
Future wireless networks are expected to support a wide variety of sometimes conflicting requirements.
6G is expected to be the first wireless standard requiring hyper-fast links with the per link peak throughput
exceeding the terabit per second (Tbps) mark. 6G use cases such as wireless factory automation will require
very sophisticated operations such as communication with ultra-high reliability and ultra-low latency, high-
resolution localization (at the centimetre level), and high-accuracy inter-device synchronicity (within 1 µs).
e 6G reliability and latency requirements are expected to be diverse and use case specific. One of the
most extreme is industrial control where only one erroneous bit in a billion transmied bits with a 0.1 ms
latency is permied.
We can anticipate that the data traffic and the numbers of connected things will increase substantially for
6G. Device density may grow to hundred(s) of devices per cubic meter. is poses stringent requirements
on area or spatial spectral efficiency and the required frequency bands for connectivity.
Latency
Jitter
Link Budget KPIs
(Global) Extended Range/
Coverage (Incl. Satellite) KPIs
3D-mapping Fidelity KPIs
Existing Tuned 5G KPIs
(Incl. Mobile Broadband)
Position Accuracy and
Update Rate
Cost KPIs
Energy KPIs
Inclusion of Vertical
Players in Definition of
Requirements and
Standardization
Transparency KPIs
(e.g. Related to AI)
Privacy/Security/Trust KPIs
Global Use Case
Oriented APIs
UN SDG Inspired KPIs
Open Source Everything
Ethics KPIs
TECHNOLOGY AND
PRODUCTIVITY DRIVEN KPIS
SUSTAINABILITY AND
SOCIETAL DRIVEN KPIS
16hps://www.pwc.com/m1/en/publications/documents/delivering-sustainable-development-goals.pdf.
17hps://www.huawei.com/minisite/gci/assets/files/Huawei_2018_SDG_report_en.pdf.
16 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
Security, privacy and reliability are important emerging KPIs. 6G will need to be hyper-secure with demanding
requirements for industrial and high-end users, while simultaneously being low cost and of low complexity
for Internet of ings (IoT) applications.
For future networks, wider radio bandwidths will be needed but can be found only at sub-THz and THz
bands. e utilization of that spectrum provides many challenges, but also opportunities. erefore, radio
hardware research will primarily focus on this spectrum area although 6G will also utilize all existing and
future bands at lower frequencies as enablers for mobile cellular large-area coverage. Super-efficient,
short-range connectivity solutions will be key for the 6G, an area in which the higher-frequency bands can
play a role in the future. e molecular absorption has a substantial impact on the path loss especially at
longer distances (~1…10 dB/km at frequencies up to 400 GHz)18. However, for local connectivity the impact
is still small compared to the free space loss and the THz radio spectrum can be divided into favourable
spectrum windows between atmospheric absorption peaks above 500 GHz (Figure 9). Penetration through
various materials and reflections from surfaces are other factors to be considered when categorizing the
radio spectrum in addition to technological boundaries.
Figure 9. Spectral windows, the effect of free space loss and water vapor absorption at a distance of 10 m.
e basic properties of frequency bands available for 5G and 6G are given in Table 1. One should note that
the increase in free space loss is quite small when moving into the THz region from 30 GHz onwards. If the
antenna area is kept constant, the free space loss is compensated for by the increase in the antenna gain.
Rather than the free space loss, the downside of higher frequencies is increased complexity and parallelism
in RF hardware and the reduced beam width that creates problems with signal acquisition and beam
tracking in mobile applications.
NEW 6G
OPPORTUNITY
NEW 5G BANDS
5G ADVANCED
0 100
50
100
150
200
250
300
200 300 400 500 600 700 800 900 1000
0
Frequency [GHz]
Path Loss [dB]
Free Space Path Loss at 10 m Molecular Absorption Loss at 25 °C and 10 m Total Loss at 10 m
18Terranova Deliverable D3.2, “Channel and Propagation Modelling and Characterization”, project report, August 2018.
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 17
Table 1. Spectrum bands for 5G and 6G.
e utilization of the spectrum in the THz regime needs to be
arranged based on absorption and reflection properties.
Many current IoT scenarios are range, cost and baery-limited and will not scale up easily to higher frequencies.
Conversely, data-rate-intensive scenarios, such as transferring holographic videos, require bandwidth not
available even in the current mm-wave spectrum. e utilization of the spectrum in the THz regime needs
to be arranged based on absorption and reflection properties of sub-bands to optimize use and reuse for
communications and other applications. Specifically, in scenarios supporting multiple applications, the
overlap of harmonic products must be prevented by careful frequency planning. Because sensitivity in weak
signal detection is one of the key bolenecks, preventive actions should preferably be considered already in
frequency regulation, something that has not been seen across large span of bands before.
Research questions
Q1: How to assess and quantify the metrics towards UN SDG KPIs?
Q2: What is an adequate radio channel model for 6G communication applications, and is it possible to
unify the model for whole range from GHz to THz?
Q3: What are feasible bands and what technologies are needed above 100 GHz for commercial use?
Q4: What are proper metrics for data privacy and security?
Q5: What are the real needs and requirements for future spectrum allocation and related policies?
FREQUENCY BAND
WAVELENGTH
DOMINANT
PROPAGATION
MECHANISM
DOMINANT
ATTENUATION
EFFECTS
SUPPORTED
LINK
DISTANCES
TX POWER
LIMITING
FACTOR
APPROXIMATE
SYSTEM
BANDWITH
0.3-3 GHz 3-30 GHz 30-300 GHz 0.3-3 THz 3-30 THz
100-10 cm 10-1 cm 10-1 mm 1000-100 µm 100-10 µm
LOS, Reflection,
Diraction,
Scattering,
Penetration
Free Space Loss
10 km
Regulation
up to 100 MHz
LOS, Reflection,
Diraction,
Scattering
Free Space Loss
-Transmission Loss
Through Materials
High at Upper Band
1000 m
Regulation
400 (or 800) MHz
LOS, Reflection
Free Space Loss/
Molecular
Absorption
-O₂ @60 GHz
-H₂O > 24 GHz
100 m
Technology
Up to 30 GHz
LOS, Reflection
Free Space Loss/
Molecular
Absorption
-High H₂O Peaks
<10 m
Technology
Up to 300 GHz
LOS, Reflection
Free Space Loss/
Molecular
Absorption
-High H₂O Peaks
<1 m
Technology
> 100 GHz
18 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
RADIO HARDWARE PROGRESS
AND CHALLENGES
e first 5G implementations will operate in frequency bands below 6 GHz for mobile applications and in
mm-waves for fixed wireless access. e focus for new hardware technologies for 5G research has been
primarily on the adoption of new spectrum at mm-wave bands, first in the 24-40 GHz region and then
gradually moving up to 100 GHz carrier frequencies. A great deal of research is still required to enable mm-
waves for mobile users including hardware and algorithms for flexible multi-beam acquisition and tracking
in non-line-of-sight (NLOS) environments. Energy efficiency for massive multiple-input multiple-output
(MIMO) antenna implementations is still a huge challenge. Due to the higher path loss, additional antenna
gain is needed, and communication needs to utilise directive links implemented with phased arrays19. Bulk
complementary-metal-oxide-semiconductor (CMOS) and CMOS silicon-on-insulator (SOI) technologies
provide adequate performance and meet requirements for most applications using off-chip antennas.
Antenna elements are still large compared to radio frequency integrated circuits (RFIC). Silicon-germanium
bipolar CMOS (BiCMOS) is a good option especially when approaching and exceeding 100 Gbps data
rates and 100 GHz carrier frequency.20.
e role of directive transmission and reception becomes even more evident when carrier frequencies
are further increased towards 1 Tbps link speeds. At the same time, it becomes more difficult to use
CMOS transistors at frequencies well above 100 GHz. On one hand, it is still beneficial to keep exploring
the potentiality of CMOS technology to support frequencies above 100 GHz. On the other hand, new or
sometimes conventional but beer performing hardware technologies, such as silicon germanium (SiGe)
or indium phosphide (InP), allow spectrum utilization on a broader scale with an improved RF performance.
Both physical and technological boundaries of electronic hardware and the fundamental laws of propagation
will become bolenecks or, at least, they will slow down the development.
Extended spectrum towards THz enables merging communications
and new applications, such as 3D imaging and sensing.
Short wavelengths and wider available bandwidths above 100 GHz will enable increased data rates but also
angular and ranging precision not seen before for imaging and radar applications for localization, 3D imaging and
sensing. erefore, hardware needs, bounds and opportunities for ultra-high-speed low-cost communications
and advanced sensing systems should be studied together on an unprecedented scale (see Figure 10)21.
e physical space needed for radio solutions will be reduced radically as the frequency increases: an
antenna array of 1000 antennas will fit into an area of less than 4 cm2 at 250 GHz. e surface area of current
mobile devices could host several tens of thousands of antennas. is will lead to new challenges with
integrated electronics becoming quite large compared to the corresponding antennas. Large antenna
arrays, required to achieve a decent range for communications or sensing will result in extraordinarily
narrow pencil beams. ese could provide beer security by pointing the messages just at the correct
targets, however at the same time they are prone to misalignment.
19T. Tuovinen, N. Tervo and A. Pärssinen, "Analyzing 5G RF System Performance and Relation to Link Budget for Directive MIMO," in
IEEE Transactions on Antennas and Propagation, vol. 65, no. 12, pp. 6636-6645, Dec. 2017.
20P. Rodríguez-Vázquez, J. Grzyb, B. Heinemann and U. R. Pfeiffer, "A 16-QAM 100-Gb/s 1-M Wireless Link With an EVM of 17% at
230 GHz in an SiGe Technology," in IEEE Microwave and Wireless Components Leers, vol. 29, no. 4, pp. 297-299, April 2019.
21P. Hillger, J. Grzyb, R. Jain and U. R. Pfeiffer, "Terahertz Imaging and Sensing Applications with Silicon-Based Technologies," in IEEE
Transactions on Terahertz Science and Technology, vol. 9, no. 1, pp. 1-19, Jan. 2019.
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 19
Probably the biggest challenges are related to energy consumption. In low rate sensing applications there
is a need for zero-energy, self-contained baery-less solutions with energy harvesting. At the other end, the
toughest visions and unexpected applications requiring broadband processing without a doubt will call for
huge improvements in power efficiency.
Generating RF power at THz is difficult and water absorption is harmful even outside the absorption peaks.
However, such absorption can be also utilized: seeing through clothing but not the body or observing the
absorption characteristics of various gases for environmental sensing. One interesting opportunity is also
the increased reflectiveness of surfaces at shorter wavelengths. While penetration through materials will
become an even more severe issue, the possibility to utilize reflections more effectively exists.
New paradigms for transceiver architecture and computing will be
needed to achieve 1 Tbps.
e limitations imposed by the laws of physic and the relevant technologies utilizing them will also enable
the new generation of wireless technology in 6G. e speed of transistor both in analogue and digital
signal processing becomes an issue when the targeted data rates of signal processing and utilized carrier
frequencies approach the fundamental limits of mainstream and affordable technologies.
Figure 10. 6G moving towards extremes: how to keep costs and complexity appropriate at both ends?
4G
IOT
2G
3G
4G
5G
Low High
Development Eort and Challenge
Complexity
New Technology
Opportunity
New Technology
Opportunity
Technology, Architecture
and System Choices
6G Zero Energy
IoT Node 6G Super Radio
Data Rate Bandwidth
Carrier Frequency
Power Consumption
Device Cost
20 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
Success of large-scale communications has been based on CMOS and – in most demanding RF
specifications – on BiCMOS-based semiconductor technologies that have provided continuous reduction
of cost per function and increased speed both in analogue and digital processing. Is this assumption still
valid in the future? e increased speed available from smaller transistors is not easily available as the
speed of the interfaces, even inside silicon, becomes the dominant boleneck, especially in CMOS. is is
further challenged by the more limited power delivery capability of nanoscale technologies22, which leads to
increased parallelism in all phases of signal processing. Unfavourable thermal effects, low breakdown voltages
and limited baery capacity are evident obstacles on the way towards Tbps communications. However, it is
challenging to envision the full replacement of silicon technologies and all opportunities to stretch the use of
the mainstream technologies will need further research from devices to transceiver architectures.
e size of one antenna element will become small as the half-wavelength distance between array elements
will be a few hundred micrometres even on the lower THz regime – scale that enables integration of antenna
arrays inside silicon. As the size of an antenna element becomes smaller than the associated electronics,
new approaches to the transceiver architecture will be needed. To avoid tens or hundreds of thousands
parallel transceiver frontends with antenna elements, advanced lens-based systems are likely to play a
significant role.
ere are opportunities for semiconductors, optics and new
materials in THz applications.
Material properties and unwanted parasitic effects typically get worse with increased frequency. erefore,
a lot of focus is currently on silicon-germanium heterojunction bipolar transistors (HBT) that outperforms
CMOS. Additionally, faster III-V semiconductor technologies, for example indium phosphide, deserve more
aention. e challenge of packaging and integration of various technologies from lens to digital technology
is one of the key research questions. Photonics, a dominant technology in upper THz regime and solution
for the highest speed interfaces, is a viable technology for 6G. As the so-called THz gap is narrowing down,
electronics and optics bring complementary opportunities both for very high-speed interfaces and for
visible light communications. is is one opportunity in the 6G context for short range links with specific
but inexpensive optical components and system solutions23.
Open-source platforms – a dream or a must to make the next
generation hardware and soware solutions happen?
It is evident that the complexity both in RF transceivers and in digital signal processing will need to increase
substantially from 5G to achieve the vision. A highly relevant question is, therefore, how we can achieve these
goals and demonstrate the advanced capabilities not only as stand-alone solutions but also as complete
systems. is calls for open-source platforms that enable low-level algorithmic development, and possibly
go much deeper into specific technologies than any open-source soware or hardware has done before.
22Hua Wang, Fei Wang, Huy ong Nguyen, Sensen Li, Tzu-Yuan Huang, Amr S. Ahmed, Michael Edward Duffy Smith, Naga Sasikanth
Mannem, and Jeongseok Lee, "Power Amplifiers Performance Survey 2000-Present", hps://gems.ece.gatech.edu/PA_survey.html.
23T. Kawanishi, "THz and Photonic Seamless Communications," in Journal of Lightwave Technology, vol. 37, no. 7, pp. 1671-1679,
April, 2019.
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 21
Research questions
Q1: How can electrical and optical technologies merge and specialize for different applications around
the so called ‘THz gap’?
Q2: Will silicon-based technologies perform well in THz/Tbps systems and what other technologies are
needed?
Q3: How can sufficient output power and steerable antenna arrays for communications and sensing
beyond 10 m be implemented at frequencies well above 100 GHz?
Q4: Can tunable antennas and other RF solutions be implemented at frequencies above 100 GHz and can
machine learning help in this problem?
Q5: How can the mutual requirements of communications, sensing, substance detection and imaging
coexist in the THz region?
22 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
PHYSICAL LAYER AND WIRELESS SYSTEM
No single solution addresses the needs of all vertical applications. e hugely varying system requirements,
such as massive broadband, ultra-reliable low-latency communication (URLLC), massive machine-type
communication (mMTC) and extreme power efficiency, mean many solutions will be required. Case-by-
case system optimization will be needed and compatibility across different use cases must be redefined.
e current 5G new radio (NR) network is not yet capable of meeting all the demanding design needs of
existing and emerging URLLC requirements, such as ultra-high reliability, ultra-low latency, ultra-secure
networks. Motivated by this, we examine the prospects for future physical layers and wireless systems. In
addition to the terrestrial networks, also infrastructures based on satellite and unmanned aerial vehicles (or
similar aerial platforms) will be needed to support the coverage and capacity requirements.
e energy and power consumption become particularly challenging when combined with the data
explosion and the fact that more and more data is packed and processed in tiny devices. At the same time,
also the complexity of transceiver processing and end-user applications may increase lead to excessive
energy consumption without careful design over all layers for energy efficiency24.
Artificial intelligence will play a major role both in link and system
level solutions of 6G wireless networks.
Meeting all the challenging requirements identified requires a hyper-flexible network with configurable
radios. AI and machine learning will be used in concert with radio sensing and positioning to learn about the
static and dynamic components of the radio environment. is will be used to predict link loss events at high
frequencies, to proactively decide on optimal handover instances in dense city networks and to determine
optimal radio resource allocations for base stations and users, just to give some examples. e future
wireless networks must be able to seamlessly interface with terrestrial, satellite and airborne networks.
Visible light communication is foreseen as a key enabler to achieve Tbps data rates in indoor scenarios.
New air interface enablers are needed and must be developed to meet these requirements (see Figure 11).
Most require the extensive use of ML and AI algorithms to improve the time varying performance of the air
interface. e concept of semantic communications (using the meaning of the messages for making the
connectivity and networking more efficient), is an important emerging area of research which is closely
connected to semantic AI. An important question is whether AI could be used to design optimal air interfaces
on the fly for a given environment and set of specific requirements. is suggests AI inspired air interfaces.
However, their true performance, in particular, power and energy efficiency in real use cases is an open
research problem.
24E. C. Strinati, S. Barbarossa, J. L. Gonzalez-Jimenez, D. Ktnas, N. Cassiau, and C. Dehos, “6G: e next frontier,” arXiv:1901.03239
[cs], Jan. 2019.
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 23
Figure 11. Challenges of future wireless.
New grant-free access methods are critical for truly massive
machine-type communication.
Extending a trend that began in 5G, 6G systems will have to flexibly enable mMTC use cases, supporting
a massive number of low-power and low-complexity devices while aaining high spectral efficiency. Such
requirements are especially demanding for the Internet of ings, where devices sporadically generate
short packets and the overhead for resource allocation may outweigh the actual exchange of information.
Modern random-access solutions for data delivery based on proper protocol design and leaning on
successive interference cancellation25, 26 may become a key enabler in this direction. Indeed, already adopted
in some satellite standards, they have been shown to approach the performance of scheduled access while
implementing a truly grant-free approach. Furthermore, modern random-access protocols leverage the
joint design of the physical and the MAC layers to boost the achievable throughput. ese may prove to be
useful for short-range connectivity solutions, i.e., in non-cellular domain.
To fully profit from such a tight integration, optimizations of the data frame structure as well as the forward
error correction design should be considered. Aention must be devoted to the choice of the modulation
scheme, which needs to be robust with respect to limited knowledge of the channel state, and on the
extension of the 5G channel coding options towards short, low-rate packets.
Orthogonal
Multiple Access
Wireless Energy Transfer
Visible Light Communication
THz Communication
IoT Nodes
Meta Materials Based
Large Intelligent Surface
INDOOR COMMUNICATION
Large Antenna Arrays
for Massive MIMO
Frequency, Time, Code
Frequency, Time, Code
25M. Berioli, G. Cocco, G. Liva, and A. Munari, “Modern Random Access Protocols,” NOW Publisher, 2016.
26F. Clazzer, A. Munari, G. Liva, F. Lazaro, C. Stefanovic, and P. Popovski, “From 5G to 6G: Has the Time for Modern Random Access
Come?” in Proc. 6G Wireless Summit, Levi (Finland), Mar. 2019.
24 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
Signal shaping is a way to achieve record-high spectral efficiency.
Achieving enhanced performance in bit rates will require the use of very high constellation modulation.
However, these high order constellations are sensitive to non-linearities in the transmission medium. Signal
shaping for quadrature amplitude modulation (QAM) may be able to overcome some of these challenges.
Signal shaping comes in two varieties: geometric and probabilistic. Both geometric and probabilistic QAM
constellation shaping are expected to achieve record-high bits/s/Hz/polarization in optical and THz
wireless communication systems.
Analogue modulation schemes in 6G?
Orthogonal frequency-division multiplexing (OFDM) has proven to be very efficient for broadband
connectivity. It has been proposed earlier also for multiband OFDM versions of ultra-wide-band (UWB)
systems at 60 GHz with greater than 500 MHz bandwidths. When the transmission bandwidths are taken
to extremes, such as several or even tens of GHz at a few hundreds of the GHz spectrum band, the classical
transceiver design starts to fail, and multicarrier modulations do not work as with current technologies.
More robust analogue modulation schemes will be needed instead.
e strongest security protection may be achieved at the physical layer.
Future optical wireless communication could rely on quantum key distribution (QKD) schemes that
could provide some unique physical layer security features, thereby enabling the required hyper-secured
networks for certain 6G applications and use cases. QKD provides a secure way of distributing secret keys
between two users. In this way, the secrecy is ensured through quantum mechanics instead of complex
computation. Furthermore, authentication by a physical layer signature, such as RF fingerprinting, and
some other technologies, such as randomization of MIMO transmission coefficients, coding, etc., could
potentially be used in 6G.
Backscaer communications using RF power for connectivity and
computation may enable hyper-low-power communications.
e overall network and system-level energy efficiency specifically, the number of bits per joule, needs to
improve significantly to support the requirements of 6G. is requires the optimization of the radio resources
so that a controlled balance between the transmit energy and required processing energy is systematically
designed. e approach calls for coding, modulation, transmit and receive processing together with
power and frequency allocation in an energy-efficient manner. Furthermore, hyper-low energy (sub mW)
capabilities are needed in terminals and even more in low-power IoT nodes. Most of this can be achieved by
appropriate RF and baseband hardware design, but low-power coding, modulation and physical layer (non-
OFDM) also needs to be addressed. Backscaer communications and energy harvesting from environment
and RF waveforms will also enable the long lifetime of IoT nodes with non-replaceable baeries. Additionally,
backscaer communications using RF power for connectivity and computation may provide a pathway
towards hyper-low-power communications.
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 25
6G wireless networks may shape the radio environment to their liking.
Driven by a revolution in electromagnetically tunable surfaces (e.g., based on metamaterials), 6G will control
signal reflections and refractions from large intelligent surfaces (LIS). Open research problems range from
the optimized deployment of passive reflectors and metamaterial-coated smart surfaces to AI-powered
operation of reconfigurable LIS. Fundamental analysis to understand the performance of LIS and smart
surfaces, in terms of the rate, latency, reliability, and coverage is needed. Another important research
direction is environmental AI whereby smart surfaces learn and autonomously reconfigure their material
parameters. Challenges include how to focus signals with different angles of incidence in large meta-
material surfaces require controllability of reflection/refraction coefficients. ML-driven smart surfaces in
mobile environments may require continuous retraining, in which the access to sufficient training data, high
computational capabilities, and guaranteed low training convergence are needed.
Holographic radio could be made possible with 6G by using LIS and similar structures. Holographic RF
allows the control of the entire physical space and the full closed loop of the electromagnetic field through
spatial spectral holography and spatial wave field synthesis. is would greatly improve the spectrum
efficiency and network capacity and would help the integration of imaging and wireless communication.
Research Questions
Q1: How to design channel coding, modulation, detection and decoding with high rates, low latency, high
reliability and large bandwidths?
Q2: How to decode Tbit/s communications (speed)? What kind of constellation shaping is needed?
Q3: How to design the systems to satisfy high-energy-efficiency and low-cost requirements? How will we
enable true baery-less operations?
Q4: How to enhance information security, privacy and reliability via the physical layer technologies? Can
quantum key distribution with optical (or microwave further in future) be practical?
Q5: How to design mm-wave/THz links, systems and transceivers efficiently? How will it be possible
to compensate for or sustain phase noise? What are the roles of coherent, noncoherent, partially
coherent systems? How will it be possible to realize mobile positioning, channel acquisition
and tracking?
Q6: How to carry out massive MIMO and smart beam steering with active antenna arrays merged with
lens antennas? How to design systems with large intelligent surfaces?
Q7: How to design efficient interfaces between high performance computation platforms and RF chains?
Q8: How to cope with very high speeds of trains and drones to support the connectivity? Can and should
we still use multicarrier technology? Do we need new waveforms such as those based on special
affine Fourier transform?
26 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
6G NETWORKING
6G needs a network with embedded trust.
By 2030, the digital and physical worlds will be deeply entangled, and people’s lives will depend on reliable
operation of the network. Major industrial value will be lost if networks fail. Whereas in the digital world
an aack may compromise intangible assets, in the cyber-physical world physical assets could be stolen,
incapacitated or harmed by digital aacks. Malicious cyber activity could lead to loss of property and life.
To counteract, we should embed ubiquitous trust model into the network27, so that users can trust the
communication over the network. Users here mean both individual and organizational entities. e trust
model should ubiquitously collect evidence of misbehaviour and provide indirect reciprocity and non-
repudiation of actions. For safety, trust and security-critical services, the network should provide embedded
distributed denial of service (DDoS) mitigation and protection from other aacks with fast and accurate
trace back to the resources used for the aack, as well as automated means of pushing the mitigation back
towards the aacker. Devices should be able to see only expected traffic while non-expected traffic should
be dropped by the network.
6G adopts the idea of ID/locator split and mainly relies on private
addresses for devices.
Embedding trust into the network needs a more stable ID for devices and nodes than just an address
that may be translated or dynamic. Each device should have at least a unique name or several names that
the network will be able to translate into addresses as well as back to IDs as needed. Devices should be
allocated either just a private address, or like classical IP hosts, they may have a globally unique address.
When aached, the device should be able to control its own reachability. A natural consequence of the
addressing principle is that end-to-end communication “layer” is separated from packet forwarding for the
users. Like Soware Defined OpenFlow networks, the network can use several forwarding protocols, such
as IPv4, IPv6, Ethernet, and several tunnelling protocols.
is is illustrated in Figure 12 and described further as follows. End-to-end network connectivity is from
one customer network to another across a wide area. Technology choices in each area are independent
due to the generic flow abstraction created by the edge node. e end-to-end connectivity layer manages
the willingness to communicate on top of the heterogeneous packet forwarding layer. e edge node
has a registry of served hosts. It assigns and maintains stable IDs for all served hosts and translates IDs to
addresses and addresses to IDs on request. e edge node collects evidence of behavior of all seen entities
against the ID of the entity supporting and using the idea of reputation of the hosts and network entities.
27R. Kantola, H. Kabir, P. Loiseau, Cooperation and end-to-end in the Internet, International Journal of Communication Systems, Wiley, 02/2017.
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 27
Figure 12. Communication in 6G can be over a chain of trust.
6G will create data markets – privacy protection will be a key enabler.
Networks will generate an unprecedented amount of information about people (IoT, Industrial IoT, eHealth,
Body area networks and so on). e IIoT will generate significant amounts of business sensitive and personal
data. Internet companies have shown how lucrative the use of private information is. Private information
collected from the physical world can be very sensitive and used against people’s interests in many ways.
We believe that to make 6G acceptable to society, the protection of private information will be a key enabler
to realize its full potential.
A fair market requires that it is possible to protect business sensitive data. Users should ultimately be able to
control and manage their private data with a simple and intuitive user interface. e ownership and control
over personal data should be given to the person or the entity in question.
Some of the data generated by 6G devices and elements in both public and private networks has value for
many societal functions and possibly to other private corporations than the one that collects the data. e
6G data markets offer a natural new business case. Clear rules for this market are needed, so that all types
of actors, including ordinary consumers, can enter it.
6G needs an upgraded networking paradigm.
e current Internet paradigm is oen referred to as “best effort” delivery. To address the need for
differentiated service quality, 5G envisions slicing, in which network resources, capacity and features are
tailored to the use case by applying traffic management, linking and computer resource allocation and a
choice of virtualized network functions that control and process the traffic in the slice. A slice can offer the
best effort to its users or apply some QoS schema for processing the packets.
In 6G, the 5G paradigm will be refined and expanded. One possibility is to virtualize (critical) end-to-end
connectivity from devices through the mobile network to the packet data network and to the cloud. Under the
6G paradigm, the network seeks to maximize user welfare or Quality of Experience (QoE) by several technical
Communication Trust
Trust
IP/ID
Customer
Network
A
Flow State
Ubiquitous
Evidence
Collection
IP/ID
Customer
Network
B
Flow State
Ubiquitous
Evidence
Collection
Customer Network
Provider A
Customer Network
Provider B
28 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
means such as intelligent traffic management, edge computing, policies set by the user either proactively or per
transaction or through traffic orchestration. e laer may, for example, use policies set by the user or by the
operator for a group of subscribers who are each treated equally within that group. e network is neutral in the
sense that it treats all applications in a slice equally and all users with the same type of subscription equally.
It is possible that by the time we reach 6G the net neutrality regulations will be updated and will mandate
MNOs to offer value added security services to users under the users’ control. Such a regulation would
define reasonable and understandable liabilities for users who have been unable to take care of the security
of their own devices in case they have been used in aacks against other users. At the same time, the
network should provide a fair basis for service and application competition to maximize end user choice.
6G research will need to investigate alternative divisions of responsibilities between the private and public
networks. e seamless integration of short-range connectivity solutions with large-coverage cellular
systems will need to become more ubiquitous and given more impetus in development and standardization.
Artificial intelligence and block chains may play a major role in 6G
networks.
Recently, there has been a growing interest in machine learning (ML) and artificial intelligence. ML relies on Big
Data that is mined to gain information and knowledge. is approach is a reasonable candidate for detecting
malicious behavior of a remote entity. ere are also other needs in networking that require “intelligence” such
as self-configuration or managing complexity. Besides Big Data, AI relies on an abundance of computing
power. 6G will use the increasing computing power for coping with the higher bitrates but also for gaining
added flexibility. is will increase power consumption dramatically unless tackled in the overall system design.
Another new technology aracting high hopes is block chain also known as distributed ledgers. Without a
central authority, in a distributed manner, this technology allows storing and sharing information that does
not change too oen. e full record of the changes is also kept. is may give rise to new ways of organizing
data markets or helping to maintain trust in an inter-operator seing.
Research questions
Q1: How to define a new networking paradigm that equally supports consumers, corporate and life and
mission critical communications? What regulatory changes are needed to allow such innovation?
Q2: How to embed trust and security in the network? How can the style of isolation provided by
virtualization be leveraged to secure end to end connectivity and communications?
Q3: What kind of data market business models are feasible and what technology is needed to support
them? What will the architecture look like and how can the non-tech users easily make use of it?
Q4: What kind of network functionality, interfaces and protocols are needed to support the new splits
of responsibilities between wide area and local area players and between consumer and vertical
market players?
Q5: How can virtualization be enhanced and improved to support maximum flexibility of networks at a low
cost for both non-critical and critical communications?
Q6: What new computing and soware technologies can be leveraged in 6G?
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 29
NEW SERVICE ENABLERS
6G is not only about moving data: it will become
a framework of services, including communication services.
Up to the 5G cellular system, the focus in cellular development has been on the communications aspects,
while other services, such as positioning, have had low priority: they have been introduced quite late into the
system design. is has not led to optimal performance or the full utilization of system capabilities. Future
services, such as mixed reality, will be so difficult to produce and require so many component enablers,
such as positioning, 3D mapping, fusion of digital content with physical model, and extremely high-speed
communication at low latency, that the co-design of the necessary enablers is not only desirable – it is
critical for a good mixed reality performance. Dense, wireless networks, with high-frequency antenna arrays
and a lot of computing power at the edge, offer natural enablers for such integrated services. e challenge
is to make these happen in an energy efficient manner.
In 6G, all user-specific computation and intelligence
may move to the edge cloud.
AI is witnessing an unprecedented surge from the wireless community driven by recent breakthroughs
in deep learning, the increase in available data, and the rise of smart devices28. Imminent use cases for AI
(particularly for reinforcement learning) revolve around creating self-learning networks and systems that
can autonomously manage resources and control functions. In addition, with the proliferation of a new breed
of autonomous devices sensing, communicating, and acting within their local environments. It is inpractical
to transmit a massive amount of local data to the centralized cloud for training and inference. is calls for
new neural network architectures and their associated communication-efficient training algorithms over
wireless links, while making real-time and reliable inferences at the network edge. Such architectures also
pose new challenges: limited access to training data, low inference accuracy, lack of generalization, and
limitations of processing power and memory for edge devices.
Edge computing29 opens new possibilities for running computation intensive, low-latency user applications
on the infrastructure side (Figure 13). An example of such computations would be foveated rendering
for mobile virtual reality experiences30. Another example would be fusion of the physical and digital world
(matching virtual content with 3D point cloud) for mixed reality applications (Figure 14). A third interesting
possibility is a local and instant information service: the edge cloud could provide fast discovery of people,
services, devices, resources and any dynamic and highly local information near the user that cannot be
collected by centralized search engines. Such an edge information service platform could be used in the
creation of a local and dynamic marketplace for services, things and information. Extreme cases for edge
computation would be a thin user client, essentially a light low-energy device capable of interacting with
human senses or neural system, with all the user specific computing occurring in edge cloud.
28Z. Zhou, X. Chen, E. Li, L. Zeng, K. Luo and J. Zhang, "Edge Intelligence: Paving the Last Mile of Artificial Intelligence with Edge
Computing," in Proceedings of the IEEE.
29N. Abbas, Y. Zhang, A. Taherkordi and T. Skeie, "Mobile Edge Computing: A Survey," in IEEE Internet of ings Journal, vol. 5, no. 1,
pp. 450-465, Feb. 2018.
30M. S. Elbamby, C. Perfecto, M. Bennis, K. Doppler, "Towards Low-Latency and Ultra-Reliable Virtual Reality," IEEE Network, special
issue in URLLC, 2018.
30 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
Figure 13. User applications and their mobility using edge technology.
Integration of sensing, imaging and highly accurate positioning
capabilities with mobility opens a myriad of new applications in 6G.
Diverging from current networks, future communication systems will become pervasive across multiple
industry verticals and thus enable a plethora of services that require positioning, such as assets tracking,
context-aware marketing, transportation and logistics systems, augmented reality and health care. In
fact, traditional localization methods relying on GPS satellites and cell multilateration are limited or even
impractical in urban and indoor deployment scenarios. Very high carrier frequencies, large bandwidths,
massive antenna arrays, densification and device to device communication are upcoming technologies
which are mostly celebrated for their communication benefits, whereas their inherent localisation potential
is typically neglected. For instance, while 3D beamforming allows for beer spectrum utilization and signal
quality overall, it also enables precise positioning for IoT applications.
RF-based sensing is another localisation opportunity enabled by high carrier frequencies of future
networks. For instance, 3D THz imaging can improve traffic safety with accurate position determination and
object detection. 3D mapping based on optical or radio technologies will be a crucial component in future
mixed reality systems and would be a natural part of the future edge services. 5G and beyond systems
supporting large bandwidths (in the order of hundreds of MHz) with massive antenna arrays would also
offer the potential to carry out high-accuracy RF positioning and tracking. However, full application of such
technologies still suffers from sufficient transmier-to-receiver isolation: it is likely that next generation
access points will be able to communicate while simultaneously resolving reflections from distinct targets.
Spectral environmental sensing at THz frequencies is another interesting new opportunity. It allows aspects
such as the detection and identification of harmful or poisonous gases in the environment.
MOBILITY MOBILITY
USER 1
USER 1
USER 2
Local Context A Local Context B Local Context C
USER 1 Application
+ User Context
USER 1 Application
in Preparation
USER 2 Applications
+ User Context
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 31
Figure 14. 6G networks can handle positioning, 3D point cloud mapping, and mixed reality data fusion in the edge.
Trust and privacy are key prerequisites for a successful
6G service platform.
e enablers discussed above will process and store personal and very sensitive information about the
users. As an example, imagine that your banking or authentication application is running on the network
edge, instead of your personal mobile device. It is inconceivable that anyone would agree to run such
applications in the network without utmost focus on trust, privacy, and information security31. is is
a somewhat different goal from the security of the communication system. e requirement for privacy
and security for the service enablers is higher as they process personal data without protection of E2E
encryption, such as a VPN (obvious if the user applications run on the edge). e edge services issue is
reminiscent of cloud services but adds the challenge of mobility since the user applications and context in
the edge need to follow the user.
Eye Tracking &
Hand Imaging
Local AR/VR
Computations
Access Link: Images,
Models, Audio, Control,
Measurements
Edge AR/VR
Computations
THz Imaging: Point
Cloud Measurements
Sounding Signals:
Position & Orientation
Measurements
Fiber/mm-Wave
Backhaul
EDGE CLOUD
6G ACCESS NODES
31P. J. Sun, "Research on the Tradeoff Between Privacy and Trust in Cloud Computing," in IEEE Access, vol. 7, pp. 10428-10441, 2019.
32 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
Research questions
Q1: How to build trust, security and privacy solutions for mobile edge services and sensitive services such
as accurate positioning?
Q2: How to reach cm-level positioning accuracy in outdoor and indoor spaces?
Q3: How to provide network-based 3D sensing/imaging capability for mixed reality services?
Q4: What user applications would benefit greatly from edge services and how?
Q5: How to provide edge services that have low latency, have access to local information and that follow
the user constantly (i.e. have mobility even across network boundaries)?
Q6: What would the role of edge AI be for service management and system orchestration? What new
requirements (e.g. stemming from privacy, security, locality or distribution) does the edge-native
environment set on current AI methods, and how can the current AI methods fulfill these
requirements?
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 33
CONTRIBUTORS
We thank all the aendees who contributed to the 6G Wireless
Summit white paper workshop:
Behnaam Aazhang, Petri Ahokangas, Hirley Alves, Mohamed-Slim Alouini, Jaap van de Beek, Howard Benn,
Mehdi Bennis, Jean Claude Belfiore, Emilio Calvanese Strinati, Fan Chen, Kapseok Chang, Federico Clazzer,
Sudhir Dixit, Kwon DongSeung, Marco Giordani, Werner Haselmayr, Jussi Haapola, Eric Hardouin, Erkki
Harjula, Jari Hulkkonen, Jari Iinai, Markku Juni, Riku Jäni, Raimo Kantola, Marcos Katz, Zaheer Khan,
Kwangseon Kim, Namseok Ko, Young Jo Ko, Joonas Kokkoniemi, Pekka Kyösti, Sandra Lagen, Mai Latva-
aho, Steven LaValle, Janne Lehtomäki, Marko Leinonen, Kari Leppänen, Calvin Li, Carlos Lima, Madhusanka
Liyanage, Lauri Lovén, Nurul Huda Mahmood, Marja Matinmikko-Blue, Marco Mezzavilla, Preben Mogensen,
Alain Mourad, Andrea Munari, Rickard Nilsson, Zhisheng Niu, Timo Ojala, Ian Opperman, Janne Peisa,
Ella Peltonen, Ullrich Pfeiffer, Pekka Pirinen, Petar Popovski, Esa Posio, Harri Posti, Ari Pouu, Ravikumar
Pragada, Aarno Pärssinen, Nandana Rajatheva, Harri Saarnisaari, Raja Sairaju, Rahim Tafazolli, Tarik Taleb,
Hugo Tullberg, Mikko Uusitalo, Harish Viswanathan, Seppo Yrjölä and Peiying Zhu.
Special thanks to co-editors:
Markku Juni, Raimo Kantola, Pekka Kyösti, Steven LaValle, Carlos Morais de Lima, Marja Matinmikko-Blue,
Timo Ojala, Ari Pouu, Aarno Pärssinen and Seppo Yrjölä.
34 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE | 35
36 | KEY DRIVERS AND RESEARCH CHALLENGES FOR 6G UBIQUITOUS WIRELESS INTELLIGENCE
6gflagship.com
6G Research Visions 1
Key Drivers and Research Challenges for 6G Ubiquitous Wireless Intelligence
Mai Latva-aho, Kari Leppänen (eds.)
6G Flagship, University of Oulu, Finland
September 2019
ISBN 978-952-62-2353-7 (print)
ISSN 2669-9621 (print)
ISBN 978-952-62-2354-4 (online)
ISSN 2669-963X (online)
