Sharing of Licensed Spectrum - a review and tutorial

Licensed Spectrum Sharing Schemes
for Mobile Operators
(concepts, challenges, SOTA approaches, and future outlook)
Wednesday, 19 October 16 1
Provided by:
Roya H. Tehrani, Prof. Klaus Moessner, Dr Seiamak Vahid
{R.Hosseinitehrani, K.Moessner, S.Vahid}@surrey.ac.uk

This work was supported in part by university of Surrey 5GIC members, in part by the EU funded
H2020 5G-PPP Project SPEED-5G under Grant 671705, and in part by the European Commission
Cognitive Radio Standardization-Initiative Project under Grant FP7-318563.

Outline
Ø  Introduction
Ø  Spectrum Authorisation Regimes
Ø  Licensed Spectrum Sharing Concepts
Ø  Survey of SOTA Approaches
Ø  Summary and Future Outlook

Introduction
v  Motivation from 5G perspective
v  Basic concepts
v  Spectrum sharing in 5G mobile communication systems
v  Pre-requisites to make spectrum sharing practically viable

Motivation from 5G perspective
q  Mobile operators are anticipating considerable increase in capacity demand to
keep pace with growth of the traffic (7.7 billion mobile broadband subscriptions and 1.5
billion cellular IoT devices by 2021)
q  The support of 10-100 times of higher end-user data rates, i.e., 10Gb/s for low
mobility and 1Gb/s for high mobility, calls for exploiting considerably wide
bandwidth
0
5
10
15
20
25
30
35
40
Permonth(ExaBytes)
Mobile Traffic growth, for various traffic types
(2015-2021)
2015
2021
By 2021 70%
of traffic will
be video

q  Licensed spectrum is currently available for mobile uses in small chunks
(contiguous 45MHz at most), which fails to meet the 5G target data-rate
q  Significant portion of sub-6GHz bands (with ideal propagation characteristics for
mobile uses), is currently allocated to non-mobile spectrum users, e.g., military,
radar, etc., across the globe
q  Re-purposing of already allocated spectrum is not viable over short time scale,
and also is costly
Spectrum *
for Mobile Uses
700MHz
(Band 20)
800MHz
(Band 20)
900MHz
(Band 8)
1800MHz
(Band 3)
1900MHz
(unpaired)
2100MHz
(Band 1)
2600MHz
(Band 7)
Maximum
Assigned
Contiguous
Bandwidth
FDD
30MHz
Uplink
and
Downlink
FDD
10 MHz
Uplink
and
Downlink
FDD
7.8MHz
Uplink
and
Downlink
FDD
45MHz
Uplink
and
Downlink
TDD
10 MHz
FDD
20MHz
Uplink
and
Downlink
FDD
35MHz
Uplink and
Downlink,
TDD 10MHz
* Source: http://www.spectrummonitoring.com/frequencies/ , UK Mobile frequencies, updated 17 Aug 2016
Motivation from 5G perspective

Basic Concepts
q  Under-utilisation of allocated bands by incumbents have been evident for some
time now
q  This will result in availability of spectrum opportunities in temporal, spatial, code,
and power dimensions
q  Spectrum opportunities can be utilised by mobile operators in a shared manner
q  Spectrum sharing allows mobile operators to acquire additional capacity in lower
costs

Amount of spectrum (in MHz) allocated to
licensed and licence exempt uses in various frequency ranges in the UK *
* Figure source: J. Burns, S. Kirtay, Ph. Marks, “Future use of Licence Exempt Radio Spectrum”, A report for the UK Spectrum Policy Forum, May 2015,
Available: http://www.plumconsulting.co.uk/pdfs/Plum_July_2015_Future_use_of_Licence_Exempt_Radio_Spectrum.pdf
Basic Concepts

*Figures source: M. Mehdawi, et al, “Spectrum Occupancy Survey In HULL-UK For Cognitive Radio Applications: Measurement & Analysis”, international
journal of scientific and technology research Vol. 2, ISSUE 4, Apr. 2013
Spectrum occupancy monitoring during 24 hours in the UK (Hull)
Received power vs. frequency (spanning 80 to 2700MHz) *
Basic Concepts

2x
3x
10~15x
10x
?x
1.5x
Spectrum Sharing
(sub-6GHz)
High Frequencies (6GHz+)
ITU WRC-15 target bands for 5G**
Cell Densification
(small cells and new cell sites)
Other opportunities?
Spectrum Refarming (sub-6GHz)
5G 1000x Capacity
Demand
Spectral Efficiency Improvement
(featuring massive MIMO, MU-MIMO, D2D, efficient air
interface, radio resource management enhancements, etc.)
*The factors in the figure are based on our estimation, more information available at:
https://www.itu.int/en/ITUD/Technology/Documents/Events2013/RegionalForumIMT_ARB_Tunis_May2013/Presentations/
RegForumIMT_2013_ARB_Tunis_May13_Presentation_FMigneret.pdf
* * : https://www.ericsson.com/mobility-report/the-need-for-spectrum-harmonization
ITU WRC-15 target bands for 5G**
Capacity Enablers in 5G Mobile Communication Systems

Figure Source: G. Ding et al., "On the limits of predictability in real-world radio spectrum state dynamics: from entropy theory to 5G spectrum sharing,"
in IEEE Communications Magazine, vol. 53, no. 7, pp. 178-183, July 2015
5G Spectrum Usage Paradigms

Wednesday, 19 October 16 12Wednesday, 19 October 16 12
Pre-requisites to Make Spectrum Sharing Practically Viable
Regulations
(sharing frameworks and authorisation regimes)
Business enablers
(e.g., pricing policies, investigation of
costs of deployment)
Technical enablers
(e.g., access techniques)

Spectrum Authorisation Regimes
v  General taxonomy
v  Licensed access schemes
v  Light licensed access schemes
v  General access schemes

General Taxonomy
q  Authorisation regimes can be distinguished based on the following attributes:
§  Level of interference protection guarantees (uni/bi lateral)
§  Level of spectrum access guarantees
§  Level of spectrum utilisation efficiency
§  License fee
*Figure Source: METIS, "Intermediate description of the spectrum needs and usage principles," Mobile and wireless communications Enablers for the Twenty-twenty Information Society,
Tech.Rep.,2013
Taxonomy of
Authorisation Regimes*
Individual Authorisation
(Licensed Access)
Dedicated
Access
Typical Bands
For Cellular
Networks
Co-Primary
Shared Access,
including Mutual
Renting and
Spectrum Pooling
Shared use of
exclusive
bands
between
operators
Licensed /
Authorised
Shared Access
(Vertical sharing)
Shared use of
exclusive bands, e.g.,
2.3-2.4 GHz between
an incumbent and
operators
Light Licensing
General Authorisation
(License Exempt
Access)
Unlicensed
Shared Access
Shared Use of
ISM bands such
as 2.4 and 5.8
GHz with
Bluetooth,
WLAN, etc.
Secondary
Horizontal
Shared Access
Shared use of
terrestrial TV
broadcasting
e.g., 470 MHz
to 790 MHz
Unlicensed
Primary Shared
Access
Shared use
of bands in
1900 MHz
with DECT
Scope of this talk

Licensed Access Schemes
Attributes
Licensed access schemes
Interference
protection
Guarantees *
Spectrum access
guarantees **
Spectrum
utilisation
efficiency
License
fee
Dedicated allocation to any
spectrum user High High Low High
Shared use of spectrum between
mobile operators; Inter-operator
spectrum sharing
Medium-High
(Scenario dependant)
Medium-High
(Scenario dependant) High Low
Shared use of spectrum between
non-mobile service providers and
mobile operators, e.g., License
Shared Access (LSA)
Medium-High Medium-High High Low
* Can be measured based on pre-defined/agreed threshold (e.g., Interference threshold for protection, and 98%QoS
or user satisfaction for access guarantees)
** The actual values are dependent on efficiency of the proposed sharing algorithms

* Fall between the individual and general authorisations in a way that based on different sharing players, it can lie either in the general or individual
authorisation regimes.
Attributes
Interference
protection
guarantees
Spectrum access
guarantees
Spectrum
utilisation
efficiency
License fee
Light Licensing
schemes *
Variable
Variable
(is agreed based on
the level of tolerable
interference)
Variable
Variable
(is defined based on
the level of
interference)
q  Allows spectrum holder to adapt its interference tolerance to exchange for
an economic compensation based on predictable interference conditions
q  Light licensing involves interference as a dynamic tradeable criteria to use
the spectrum and for pricing reasoning
Light Licensed Access Schemes

License Exempt Access Schemes
Attributes
General access schemes
Interference
protection
guarantees
Spectrum
access
guarantees
Spectrum
utilisation
efficiency
License
fee
Secondary Horizontal Shared
Access; e.g., shared TVWSs Low Low variable Low
Unlicensed Shared Access; e.g.,
Licensed Assisted Access (LAA)
LTE-Unlicensed (LTE-U)
No/low No/low variable No
Unlicensed Primary Shared
Access
No/low No/low - No

v  Which sharing scenarios are available?
v  What are the potential use cases?
v  What are the technical challenges?
v  Which techniques can be applied to deploy spectrum sharing?
v  What are the business considerations of deployment?
Licensed Spectrum Sharing Concepts

Classification of the licensed sharing schemes based on types of sharing players:
q  Homogenous sharing players: sharing players are of the same nature
(e.g., multiple mobile operators), and encompasses various types
q  Heterogeneous sharing players: sharing players are of different nature
(e.g., military service providers and mobile operators)
Which Sharing Scenarios Are Available?

Licensed Sharing Deployment
Scenarios for
Mobile Cellular Networks
Inter-operator Spectrum/
Resource Sharing (Homogenous
sharing players)
Inter-operator
Spectrum Sharing
(No RAN Sharing)
Mutual
Renting
Spectrum
Pooling
Inter-operator
Spectrum and RAN
Sharing
Inter-operator
National Roaming
Multi-operator
RAN and Spectrum
Sharing
Spectrum Sharing between
Operators and Incumbent
(Heterogeneous sharing players)
Licensed Shared Access (LSA)
Authorised Shared Access (ASA)
Spectrum Access System (SAS)
Classification of Licensed Sharing Schemes

q  From property sharing perspective
§  Only the spectrum is shared
§  Both network infrastructure/assets and the spectrum are shared
q  From network deployment perspective
§  Non-collocated deployment of mobile operators (Cells of different operators may
partially overlap)
§  Collocated deployment of mobile operators (Cells of different operators may
completely overlap)
q  From coordination perspective
§  Real-time coordination between participating operators
§  No coordination between participating operators
Homogenous Sharing Players-Sharing Perspectives

Where only the spectrum is shared
1)  Mutual renting; mobile operators rent part of their exclusive spectrum to be exploited by
other mobile operators (can be on bi-lateral or unilateral basis)
2)  Spectrum pooling; mobile operators can exploit licensed bands in a shared manner with
the same right of access (e.g., the bands are made available from LSA framework)
•  In both cases operators may be deployed collocated or non-collocated
•  Figure source: http://www.google.ch/patents/US20140206377
Inter-operator Spectrum Sharing

Where both the spectrum and/or RAN infrastructure can be shared
q  Where users are served by the host (visited) operator, e.g., various types of
inter-operator National Roaming schemes
q  Where users are served by their home operators, e.g., various types of inter-
operator Spectrum and/or Radio Access Network (RAN) sharing schemes
Inter-operator RAN/Spectrum Sharing

q  Passive elements which do not contribute to transmission of signal (e.g., sites,
mast, cooling, etc.)
q  Active elements which contribute to transmission of signal (e.g., RAN, Radio
Remote Heads(RRHs), Core Network (CN), etc. )
q  Infrastructure sharing is expected to considerably reduce the operational costs of
deployment
Infrastructure-only Sharing Types

For RAN sharing the two most commonly used solutions are known as:
q  Multi-operator Radio Access Networks (MORAN)
§  With MORAN everything in the RAN except the spectrum is shared between multiple
operators
q  Multi-Operator Core Network (MOCN)
§  With MOCN everything in the RAN as well as spectrum can be shared between
multiple operators
With both solutions the operators can keep their existing core networks separate

3GPP.“Service aspects and requirements for network sharing,”, 3GPP, Sophia AnLpolis, France, TR 22.951 V12.0.0, Oct. 2014.
3GPP. “Network sharing; architecture and funcLonal descripLon,” 3GPP, Sophia AnLpolis, France, TS 23.251 V13.1.0, Mar. 2015.
3GPP. “Study on radio access network (RAN) sharing enhancements,” 3GPP, Sophia AnLpolis, France, TR 22.852 V13.1.0, Sep. 2014.

Inter-operator RAN and/or Spectrum Sharing

MORAN and MOCN Architecture
Figure source: http://www.unwiredinsight.com/2013/3gpp-lte-ran-sharing-enhancements

Various Types of Infrastructure-Spectrum Sharing between Operators
Figure source: http://www.analysysmason.com/About-Us/News/Newsletter/Active-RAN-sharing-Oct2014/
Passive Sharing Active Sharing
Site sharing Backhaul sharing MORAN MOCN
A
A
A
shared
B
B
B
A BA
A
B
B
A B
BABA BA
shared
shared
shared
shared
shared
shared
shared
sharedshared
Core
Network
eNodeB
Tower/
Antenna
Spectrum
Backhaul
Operator A Operator B

Comparison of MORAN and MOCN *
* Network Consolidation Cooperation for Business Success, WhitePaper,Available:http://www.huawei.com/ucmf/groups/public/documents/attachments/hw_454338.pdf
Types
Attributes
MORAN MOCN
Spectrum Exclusive Shared
Feature implementation and
configurations upgrade
Independent Shared
Operational data monitoring Independent Can be shared
Transmission pipes Independent or shared Independent or shared
Physical links Shared Shared
Core network Independent Independent
Use cases
For keeping some degree
of independence for
operators on RAN level
Where there is no
limitation for spectrum
sharing

q  Support of user device operation in a network other than its own home network
(within the same country) is referred to as national roaming
q  Valuable:
§  When mobile operators provide coverage in different geographical areas
§  For mitigating network outage and congestion
§  Suitable for Mobile Virtual Network Operator (MVNO) which does not own
RAN
§  For the new entrants at initial deployment steps
§  For the smaller operators who may not be able to sustain the costs of
covering a large territory with a low population density
q  Is done based on Service Level Agreement (SLA) before operating on shared
bands
q  User device measures the signal strength of the pilot signals (beacon signals) of
the neighbouring base stations and is connected to the strongest one
Inter-operator National Roaming

Inter-operator National Roaming Limitations
q  Roaming is not an effective solution where:
§  Outages affecting shared infrastructure
§  Outages affecting all providers at once
§  Congestions affecting all providers at once
q  Providers offering roaming can prioritise their own users to provide acceptable
level of QoS
q  National roaming cannot be used if the Home Location Register (HLR) of the
home network is unreachable by the visited operator
q  National roaming could negatively impact competition and may remove
incentives to invest in network infrastructure
q  Many mobile operators still oppose to National Roaming and aim to propose
solutions to avoid the technical, economic and competitive barriers of National
Roaming

1.  https://www.enisa.europa.eu/publications/national-roaming-for-resilience/at_download/fullReport
2.  Mobile Coverage in the UK: Government plans to tackle ‘mobile not-spots’ , Number CBP-07069, 22 September 2016 , available :
www.parliament.uk/commons-library | intranet.parliament.uk/commons-library

q  Currently involves following frameworks
o  Authorised Shared Access (ASA)
o  Licensed Shared Access (LSA)
o  Spectrum Access System (SAS)
q  Facilitates spectrum sharing between non-mobile service providers (called
incumbent) and mobile operators
q  Co-existence of incumbent and mobile cellular systems is managed through
identification of exclusion zones (temporal/spatial separation)
Heterogeneous Sharing Players

q  ASA a sharing framework which facilitates use of 2.3-2.4GHz (in the EU) and
3.5GHz (in the U.S.) bands, in a shared and non-interference basis for mobile
cellular systems only
q  LSA is an extension of ASA concept, and is proposed by CEPT ECC to support
sharing between various types of spectrum users and various bands
q  The reference architecture for both ASA and LSA is the same:
§  Spectrum owner (Incumbent network)
§  Licensees (mobile operators)
§  LSA controller
§  LSA repository
§  Interface for connectivity
q  LSA excludes concepts such as “opportunistic spectrum access”, “secondary
use” or “secondary service” where the applicant has no/low protection from primary
user(s)
Authorised/Licensed Shared Access

Figure source: http://www.alticelabase station.com/en/projects/adel.html
LSA Reference System Architecture

q  SAS framework operates in 3.55-3.7GHz (LTE Bands 42/3), and follows three-
tier strategy
§  Incumbent
§  Primary Access (PA)
§  General Authorised Access (GAA)
q  The level of interference protection between the tiers is reduced top down. So
GAA may suffer from incumbent and PA interference
Figure Source: http://www.federatedwireless.com/solutions/sas/
Spectrum Access System

SAS Framework Architecture
Figure source: http://itg.lkn.ei.tum.de/lib/exe/fetch.php?media=termine:2015-12-11-muenchen:2015_12

ASA, LSA, and SAS Frameworks Comparison
Scheme
Attribute
ASA LSA SAS *
Incumbent types Military applications/
Wireless cameras/
Satellite
Initially the same as ASA,
To be extended into various
types of service providers
Federal and Fixed
Satellite Service
(FSS)
Prospective Licensees
Mobile broadband
Service
Initially for mobile broadband
service
(to be extended for any
spectrum user)
PA: hospitals, public
safety entities,
Mobile Broadband
GAA: LAA, Wi-Fi
Spectrum range 2300 to 2400 MHz,
and 3800GHz
Not only limited to IMT bands 3550 to 3650MHz
Spectrum opportunity
detection method
Geolocation Database
(& sensing if necessary)
Geolocation Database
(& sensing if necessary)
Sensing & Database
Incumbent protection Protection from
harmful interference
Protection from
Protection from
Licensee QoS
provisioning
Predictable QoS for all
authorised users
(Must ensure certain
access guarantees and
interference protection)
Predictable QoS for all
authorised users
(Must ensure certain access
guarantees and interference
protection)
“Some” interference
protection for PA
And “no” protection
guarantees for GAA
users
* https://www.fcc.gov/document/35-ghz-sas-and-esc-application-ex-parte-status

q  Sub-urban/urban not-spot coverage enhancement
§  Setting up new sites for sub-urban, or small cells in urban scenarios are subject to
considerable costs with no considerable revenue (Figure below)
§  National roaming is an indicative solution, but not always favourable for competitive
operators
§  More spectrum with low penetration losses is required (sub-1GHz bands such as
700-to-900MHz) that can be made available through various licensed sharing
schemes
* Figure source: http://blog.3g4g.co.uk/2012/07/fundamentals-of-mobile-network-sharing.html
Potential Use Cases of Licensed Sharing Schemes

q  Urban hot-spot capacity enhancement
§  Wi-Fi offloading to handle traffic peaks in special events currently provides
unpredictable QoS
§  Licensed spectrum sharing schemes can be applied when more capacity
(and wider bandwidth) of licensed spectrum with predictable QoS is
required in hot-spots


q  Mass deployment of small cells on “shared” (licensed and/or unlicensed) bands
§  Co-existence of macro, pico and femto cells on the same cellular bands
raises the problem of interference
§  Higher frequency ranges, are made available through LSA, facilitate
deployment of small cells (mainly indoor) with low transmission power and
free of interference

q  Radio Access Technology (RAT)-specific band sharing
§  3GPP RATs such as; 2G, 3G, 4G/LTE, and LTE-A operate on different
frequency bands
§  Not all mobile operators exclusively own RAT-specific bands (known as
partial/operator-specific not-spot)
§  Inter-operator spectrum sharing schemes facilitate access to the bands in a
shared manner with lower license fees

q  Capacity enhancement considering Frequency Division Duplex (FDD) and Time
Division Duplex (TDD) band sharing
§  Feasibility of aggregation/joint-use of FDD and TDD bands in intra-
operator case has been investigated in 3GPP Rel. 12
§  This possibility can allow mobile operators to share their bands
regardless of access mode, for capacity and bandwidth enhancements

q  Spectrum sharing schemes require robust techniques for spectrum opportunity
detection
q  Due to user mobility, combined real-time spectrum opportunity detection and location
estimation is preferable but is challenging
q  Licensed spectrum sharing players require mechanisms to protect sharing players
from harmful interference
•  This calls for coordination, robust access methods between sharing players
q  The type and level of information to be exchange for the coordination might be
problematic in terms of business risks and control overheads to the existing systems
Technical Constraints and Challenges

Current CoordinaLon
Techniques in licensed
sharing
Centralised
Database Driven
approaches
Spectrum Broker
/Meta Spectrum
Scheduler/management
enLty
Distributed
Spectrum sensing
(wide ranges, e.g.,
energy detecLon, etc.)
Coordinated
beamforming
Coordinated Game-
theory based
approaches
Current Approaches to Coordination

q  Centralised techniques (Concept)
§ Database driven approaches
o  Geo-location database is an indicative example
o  It stores and processes geo-localised information of spectrum availability
o  Provides interference maps of network based on offline theoretical
propagation models
§ Meta spectrum-scheduler/Spectrum broker
o  A centralised management entity which may run various policies:
•  Allocation of shared bands based on channel quality of users
•  Auction based spectrum allocation
•  Sharing as a last resort (only when additional spectrum is required)
•  Always connected to the nearest base station (regardless of the
operator)

Centralised sharing management
entity, e.g., database, broker, etc.
Operator A
S1-MME
SI-U
Operator B
S1-MME
SI-U
An interface
with reasonable speed
MME MME
q  Centralised techniques (example architecture)

q  Centralised techniques (pros/cons)
ü  Provide accurate information regarding spectrum availability across the
network
ü  Provide reliable interference protection for sharing players
ü  Can be implemented and managed by an unbiased third-party for fair spectrum
allocation among sharing players
!  Too complex for real-time spectrum opportunity detection
!  Requires additional infrastructure such as backhaul for deployment
!  Scalability of such mechanisms for wide area deployment matters
!  Requires a third party to manage the sharing procedure
!  Imposes excess signalling overhead to the network/participating systems
!  Vulnerable to jamming attacks

q  Distributed techniques
§ Spectrum sensing
§ Coordinated beamforming
§ Game-theory based coordination

q  Spectrum sensing (concepts)
§  Any sensing capable devices can detect the presence of other devices
operating on the same bands, to avoid interference in a distributed manner
§  A wide range of sensing techniques are available, energy detection, matched
filtering, spectral correlation (Cyclostationarity), feature detection of co-existence
beacons, etc.
Figure source: T. Yucek and H. Arslan, “A survey of spectrum sensing algorithms for cognitive radio applications,” IEEE Commun. Surveys Tuts., vol. 11, no.
1, pp. 116–130, 1st Quart. 2009.

q  Spectrum sensing (Implications)
ü  Is capable for on-demand and real-time spectrum opportunity detection
ü  Except sensing capabilities, no additional infrastructure is required
ü  Only target user devices involved to perform sensing, thus, lower signalling is
imposed to the network
!  Vulnerable to some issues such as hidden node, maintenance of false alarm
and detection probability constraints
!  Not reliable for QoS sensitive services when sensing is purely performed by
user device

q  Coordinated beamforming (concept)
§  Enables mobile operators to adjust size and position of the cells to
better serve users
§  Flexibly modifying the phase and amplitude of the signals to shape and
steer the direction of the radiated beam vertically and horizontally to
create constructive or destructive interference
o  Constructive interference is used to amplify the beam in a
given direction
o  Destructive interference is used to focus the beam, enabling it
to be steered precisely
§  In the context of spectrum sharing facilitates co-existing multi-
technology deployments

q  Coordinated beamforming (concept)
Exclusive spectrum
Operator A
Exclusive spectrum
Operator B
Shared spectrum pool
Operator A
Operator B
Interfering signal
Information
X2, or any interface with reasonable speed for
inter-operator inter-RAN coordination

q  Coordinated beamforming (Implications in inter-operator spectrum sharing)
ü  Simultaneous utilisation of spectrum by multiple mobile operators in the same area
ü  Increased spectrum utilisation efficiency
!  Requires Channel State Information (CSI) sharing between sharing players
!  Requires interface (such as wired backhaul, X2, etc.) between sharing players

q  Game-Theory based coordination (Concept)
§  A well-defined technique for studying distributed decision-making in multi-user
systems
§  Applied to the problems such as; power control, spectrum allocation, call
admission control, and routing
§  Game can be cooperative or non-cooperative
§  Without coordination among users/systems, the outcomes is analysed through so-
called Nash Equilibria (NE)
§  To achieve better payoffs, cooperation between users is carried out subject to
sharing some information
§  If there are extra utilities, players may bargain (Nash Bargaining (NB)) with each
other to decide how to share the information

q  Game-Theory based coordination (Implication in inter-operator spectrum sharing)
ü  Low to no information sharing between sharing players during sharing procedure (is
mainly based on the pre-agreement)
ü  Low to no overhead is imposed to the existing networks
!  Implementation complexities
!  Uniqueness complexities
!  Low fairness guarantees between sharing players

q  Additional infrastructure
§  Depending on the information exchanged among sharing players from
static to real-time, suitable interface (wired/wireless backhaul, X2, fibre) is
required
§  Database with high computational and memory capabilities to store and
process spectrum availability information
q  Multi-band operational capabilities of devices such as user equipment and
Base station
§  Both Base Stations and User devices require software/hardware
update to be capable of operating in sub-6GHz bands (or even
6GHz+)
§  Base stations need to be capable of operating in wider bandwidths
and serving more users, thus more power in required
Business Considerations: Costs of Deployment

q  Uncertainty and business risk which makes spectrum sharing less attractive
§  Established mobile operators may realise spectrum sharing as a threat in the
market
§  Concerns of greedy use of shared bands
§  Mobile operators must be assured the license is transferred for short time
§  Mobile operators or incumbents may have concerns on sharing their private
information (e.g., real-time spectrum usage)
q  Licensing policies
§  Channel-quality based pricing
§  Game-theoretic based pricing
§  Demand-supply model in which shared bands are assigned to the highest
bidders
Business Considerations: Costs of Deployment

Survey of SOTA Approaches for Various Sharing Schemes

Technique and Mechanism Impacts
Centralised
Coordination via a third-party
(Cell level negotiation)
Sharing is performed only when
additional resource is needed and
available
+ Shows gain in terms of reduced packet drop rate compared to no
sharing
- Requires home base station to have knowledge of spectrum usage in
its adjacent cells to avoid interference
- This problem reduces gain and requires connectivity among adjacent
base stations for information acquisition
Distributed
Collaborative spectrum sensing
Energy detection by sensors
Base stations are connected with
sensors
+ Shows gains in terms of reducing packet drop rate compared to no
sharing
+ Cost of deployment can be shared among operators
- Requires backhaul to connect sensors and base stations
- Vulnerable to sensing related issues in indoor and mountainous areas
Distributed
Spectrum sensing
Energy detection by user device
Information is sent to Base station
+ Except sensing capable devices, no additional infrastructure is
required
+ Real-time spectrum opportunity detection
- Vulnerable to sensing related issues such as false alarm and
detection, hidden node problem
- Short time scale sharing (sub-frame level) requires synchronisation of
operators
Mutual Renting Approaches-SOTA

Spectrum Pooling-SOTA
Technique and Mechanism Impacts
Centralised super scheduler
allocates shared bands
Decision is made based on the
CSI of the user devices regardless
of their home operator
+ 20% increased cell sum capacity (upper bound)
- Fairness is not guaranteed among user devices of different
operators
- Requires real-time interaction between base stations and super
scheduler per user resource allocation (signalling overhead)
Coordinated beamforming
+ Increased spectrum utilisation efficiency compared to no sharing
- Requires sharing of CSI between operators
- Requires interconnection among base stations of operators
- Beneficial for only users with high SINR, close to their serving base
stations
Game-theory based approach
Cooperative games performs
based on pre-sharing agreements
among operators
+ No need for real-time inter-operators information sharing
- Efficient and fair policies are complex to implement

Technique Impacts
Centralised
Radio Network Controller is shared
between operators
(in both collocated and non-
collocated deployment)
+ Roughly 32% increase in cell capacity compared to non-sharing
- Low/no gain in the cases of symmetric traffic
Distributed
User devices sense reference signal
of host base station
No additional infrastructure is
required
+ 10% improvement in cell throughput compared to the case of
non-sharing
- Low/no gains in cases of symmetric traffic
- Increased delay, due to inter-operator handover messaging
procedure
Inter-operator National Roaming-SOTA

Technique Impacts
Multi-operator Virtual RAN, and
Spectrum Sharing
+ Enables significant reduction in capital expenditure in low
traffic areas
+ Facilitates spectrum sharing procedure among the operators
(not additional interface between participating operators is
needed)
- Requires virtualisation capable infrastructure
Inter-operator Virtual RAN Sharing-SOTA

Licensed Shared Access-SOTA
Project/Paper Incorporated technique Objective Impacts
LSA trial
demonstration
SON is integrated in LSA
controller and incumbent
user movement tracking
Reduction of delay in LSA
band-evacuation phase, and
a more robust incumbent
interference protection.
Delay reduced to 85%,
from 21s (former trials)
to 3s, and a 18%
capacity improvement
“Optimisation of
Authorised/Licensed
Shared access”
Power adaptation and
beam-steering in LTE
network
To protect incumbent users
from interference while
incorporating 2300 MHz
bands for LTE use.
30% improvement in
a v e r a g e . u s e r
throughput outside of
the exclusion zone
(where incumbent
users do not exist), and
10% improvement in
a v e r a g e u s e r
throughput within the
exclusion zones, with
power reduction and
downtilt.
“RED
Technologies” ,
“ADEL”
Radio Environment
mapping
More dynamic and accurate
s p e c t r u m o p p o r t u n i t y
detection.
Project ongoing.

Summary & Conclusion
q  We reviewed relevant 5G objectives
q  We studied various use cases in which licensed spectrum sharing can be
beneficial
q  We studied various available spectrum sharing & access techniques and their
implications on various licensed sharing schemes
q  Based on this study it can be concluded that:
§  There is still room for further improvements of current access techniques
§  Also a need for advanced novel sharing schemes

Future Outlook
q  Rooms for further improvement
§  Enhanced Inter-Operator Coordinated Beamforming Techniques
§  LSA Framework Enhancement
§  Enhanced Radio Environment Mapping techniques
§  Enhanced RAN Sharing Schemes
§  Enhanced Spectrum-Sensing Techniques
§  Inter-Operator Inter-cell Interference Coordination

Backup Slides

Regulatory Bodies Point of View of Spectrum Sharing
q  Ofcom (in the U.K.)
§  Considers spectrum sharing among federal spectrum users, such as military,
defence, etc., and mobile operators (under LSA scheme)
§  Data off-loading to Wi-Fi for indoor capacity improvement and for outdoor
MTC applications with lower QoS expectation
q  FCC (in the U.S)
§  Considers spectrum sharing among federal spectrum users such as radar,
and mobile operators (under SAS scheme)

Standardisation Activities
q  ETSI
§  Investigates various spectrum sharing schemes
§  Infrastructure sharing issues are not currently addressed
§  Applicability of cognitive techniques such as Radio Environment Maps
(REMs) are being investigated
q  3GPP
§  Increasing interest in various resource sharing scenarios, according to
identified RAN sharing scenarios, whether as a shared deployment or as a
leased asset
q  ITU-R
§  Soliciting solutions for the use of licensed “white spaces”, and licensed-
exempt bands with the aim of provisioning ubiquitous wireless
connectivity

Involved EU Projects on spectrum sharing (CORDIS) *
q  SAPHYRE
§  Provides system level analysis of inter-operator spectrum,
infrastructure/RAN, and full sharing
§  Analysis of business models, cost and advises for spectrum policy and
regulation
q  METIS I & II
§  Spectrum sharing enablers
§  Spectrum business models
§  Potential spectrum sharing technologies
q  QoSMOS
§  The initial focus was on opportunistic use of radio spectrum, for LTE
extension in TVWS
§  Developing a framework for characterisation and modelling of the radio
environment
* Other projects include: COGEU, ADEL, FARAMIR, CREW, S3ISE, QUASAR, CRS-I, CoRaSat

References
§  The information contained in this document is provided from the article
and all the references therein, which has been published. Citation
information: DOI 10.1109/COMST.2016.2583499, IEEE Communications
Surveys & Tutorials. 1553-877X ©2016 IEEE
§  If you interested to get more information, the full version of the paper can
b e d o w n l o a d e d f r o m t h e f o l l o w i n g l i n k :
http://ieeexplore.ieee.org/document/7500126/
§  Personal use is permitted, but republication/redistribution requires IEEE
permission. You may not distribute this document externally, in whole or
in part, to any other person or entity. For more information. See:
http://www.ieee.org/publications_standards/publications/rights/index.html
§  5GIC, and its contributors may have patents or pending patent
applications, publications, trademarks, copyrights or other intellectual
proprietary rights covering subject matter contained or described in this
document
Article
Full Access
Copy Right

Disclaimer
The information contained in this document is provided for tutorial purposes only and
represents the current view of ICS and its contributors, as of the date of publication. 5GIC,
and its contributors do not warrant that any approach referenced herein or any product or
technology developed in reliance upon this document, in whole or in part, will be sufficient,
accurate, reliable, and complete, free from defects or safe for its intended purpose, and
hereby disclaim all liabilities therefor. Any person making, using or selling such product or
technology does so at his or her own risk.

Sharing of Licensed Spectrum - a review and tutorial

More Related Content

What's hot (8)

Similar to Sharing of Licensed Spectrum - a review and tutorial (20)

Recently uploaded (20)

Sharing of Licensed Spectrum - a review and tutorial