Fog Computing Versus SDN/NFV and Cloud Computing in 5G

Fog Computing Versus SDN/NFV and Cloud Computing in 5G

Loading
Loading Social Plug-ins...
Language: English
Save to myLibrary Download PDF
Go to Page # Page of 105

Description: This overview is compiled, based on several public documents belonging to different authors and groups, on Wireless/4G/5G networking, Cloud Computing, SDN, NFV, Slide 2 DataSys 2016 Conference May 22, 2016, Valencia, Spain Fog Computing, etc. Conferences material, studies, research papers, standards, projects, overviews, tutorials, etc. (see specific references in the text and Reference list).

 
Author: Eugen Borcoci  | Visits: 217 | Page Views: 392
Domain:  High Tech Category: Mobile 
Upload Date:
Link Back:
Short URL: https://www.wesrch.com/electronics/pdfEL1WSS000AFJK
Loading
Loading...



px *        px *

* Default width and height in pixels. Change it to your required dimensions.

 
Contents:
Fog-computing versus SDN/NFV and Cloud
computing in 5G
Eugen Borcoci
University POLITEHNICA Bucharest (UPB)
Eugen.Borcoci@elcom.pub.ro

DataSys 2016 Conference May 22, 2016 Valencia, Spain

Fog-computing versus SDN/NFV and Cloud
computing in 5G
Acknowledgement

This overview is compiled, based on several public
documents belonging to different authors and groups, on
Wireless/4G/5G networking, Cloud Computing, SDN, NFV,
Fog Computing, etc. : conferences material, studies,
research papers, standards, projects, overviews, tutorials,
etc. (see specific references in the text and Reference list).

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 2

Fog-computing versus SDN/NFV and Cloud
computing in 5G
Motivation of this talk
Facts:
Internet and Telecom convergence → Integrated networks: Future
Internet
Novel services and communication paradigms
Content/media oriented communications, Social networks Internet
of Things, M2M and Vehicular communications, etc.
Novel, emergent technologies are changing networks and services
architectures :
Advances in wireless technologies: 4G-LTE, LTE-A, WiFi
Evolution to 5G
Support technologies


Cloud Computing



Software Defined Networks (SDN)



Network Function Virtualization (NFV)



Fog/Edge Computing

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 3

Fog-computing versus SDN/NFV and Cloud
computing in 5G
Motivation of this talk (cont’d)

Source CISCO
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 4

Fog-computing versus SDN/NFV and Cloud
computing in 5G
Motivation of this talk (cont’d)
Specific aspects in wireless/mobile environment (having impact
on 4G, 5G architectures and technologies)
the limited spectrum and bandwidth, interference
time- and location-dependent wireless link characteristics
radio resource management and allocation
mobility issues
heterogeneous RATs, terminal features
how to obtain higher capacity networks
Main topic of this tutorial :
Summary presentation of solutions to develop 5G networking
and services - based on concepts and cooperating - support
technologies as:
Cloud computing
Software Defined Networking
Network Functions Virtualization
Fog/Edge computing
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 5

CONTENTS
1. 5G Vision and Architectures
2. Software Defined Networking (SDN) and Network
Function Virtualization (NFV)
3. Cloud Computing Architectures in 5G
4. SDN and NFV in 5G
5. Fog Computing in 5G
6. Mobile Edge Computing
7. Open research topics
8. Conclusions

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 6

CONTENTS
1.
5G Vision and Architectures
2. Software Defined Networking and Network Function
Virtualization
3. Cloud Computing Architectures in 5G
4. SDN and NFV in 5G
5. Fog Computing in 5G
6. Mobile Edge Computing
7. Open research topics
8. Conclusions

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 7

1. 5G Vision and Architectures
Key Drivers, Requirements, Technologies
Driving factors for cellular network evolution 3G

4G

5G

Device, Data, and Data transfer rates
continuous growth in wireless user devices, data usage
desired: better quality of experience (QoE)
~ 50 billion connected devices will utilize the cellular network services until 2025
high increase in data traffic

Current State-of-the-art solutions are not sufficient !
Three views for 5G:
user-centric (uninterrupted connectivity and comm. services, smooth consumer
experience )
service-provider-centric (connected intelligent transportation systems, road-side
service units, sensors, and mission critical monitoring/tracking services)
network-operator-centric (scalable, energy-efficient, low-cost, uniformlymonitored, programmable, and secure communication infrastructure)

Consequence: three main 5G features
Ubiquitous connectivity: devices connected ubiquitously ; uninterrupted user
experience
Zero latency (~ few ms): for life-critical systems, real-time applications, services
with zero delay tolerance.
High-speed Gigabit connection
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 8

1. 5G Vision and Architectures
Key Drivers, Requirements, Technologies
5G disruptive capabilities
x 10 improvement in performance : capacity, latency, mobility, accuracy
of terminal location, reliability and availability.
simultaneous connection of many devices + improvement of the terminal
battery capacity life
lower energy consumption w.r.t. that today 4G networks; energy
harvesting
Better spectral efficiency
help citizens to manage their personal data, tune their exposure over
the Internet and protect their privacy
reduce service creation time and facilitate integration of various players
delivering parts of a service
built on more efficient hardware
flexible and interworking in heterogeneous environments
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 9

1. 5G Vision and Architectures
Key Drivers, Requirements, Technologies (cont’d)
Additional requirements ( and objectives) :
sustainable and scalable technology
cost reduction through human task automation and hardware
optimization
ecosystem for technical and business innovation

Application fields:
network solutions and vertical markets:








automotive, energy
food and agriculture
city management, government
education
healthcare, manufacturing
public transportation
…….

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 10

1. 5G Vision and Architectures
Key Drivers, Requirements, Technologies (cont’d)
5G - evolution of mobile broadband networks + new unique network
and service capabilities:
It will ensure user experience continuity in various situations
high mobility (e.g. in trains)
very dense or sparsely populated areas
regions covered by heterogeneous technologies
5G -key enabler for the Internet of Things, M2M
Mission critical services :
high reliability, global coverage and/or very low latency (currently they
are handled by specific networks), public safety
It will integrate: networking + computing + storage resources into
one programmable and unified infrastructure
optimized and more dynamic usage of all distributed resources
convergence of fixed, mobile and unicast/mcast/broadcast services.
support multi tenancy models, enabling players collaboration
leveraging on the characteristic of current cloud computing
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 11

1. 5G Vision and Architectures
Key Drivers, Requirements, Technologies (cont’d)
5G Key technological characteristics
Heterogeneous set of integrated air interfaces
Cellular and satellite solutions
Simultaneous use of different Radio Access Technologies (RAT)
Seamless handover between heterogeneous RANs
Ultra-dense networks with numerous small cells
Need new interference mitigation, backhauling and installation techniques.

Driven by SW
unified OS in a number of PoPs, especially at the edge of the network

To achieve the required performance, scalability and agility it will rely on
Software Defined Networking (SDN)
Network Functions Virtualization (NFV)
Mobile Edge Computing (MEC)
Fog Computing (FC)

Ease and optimize network management operations, through
cognitive features
advanced automation of operation through proper algorithms
Data Analytics and Big Data techniques -> monitor the users’ QoE

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 12

1. 5G Vision and Architectures
5G Key Requirements
Summary of 5G figures (very) for ambitious goals:
1,000 X in mobile data volume per geographical area reaching a target
≥ 10 Tb/s/km2
1,000 X in number of connected devices reaching a density ≥ 1M
terminals/km2
100 X in user data rate reaching a peak terminal data rate ≥ 10Gb/s
1/10 X in energy consumption compared to 2010
1/5 X in E2E latency reaching 5 ms for e.g. tactile Internet and radio link
latency reaching a target ≤ 1 ms for e.g. Vehicle to Vehicle
communication
1/5 X in network management OPEX
1/1,000 X in service deployment time reaching a complete deployment in
≤ 90 minutes
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 13

1. 5G Vision and Architectures
5G Generic Architecture
multi-tier arch. : small-cells, mobile small-cells, and D2D- and CRN-based comm.

Source: Panwar N., Sharma S., Singh A. K. ‘A Survey on 5G: The Next Generation of Mobile Communication’.
Accepted in Elsevier Physical Communication, 4 Nov 2015, http://arxiv.org/pdf/1511.01643v1.pdf
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 14

1. 5G Vision and Architectures
Example: Cellular systems evolution towards 5G H-CRAN
Novel proposal for 5G architecture : H-CRAN Heterogeneous Cloud Radio
Access Networks

RRH – Remote Radio Head; CoMP - coordinated multi-point; MBS Macro Base Station
Source: M. Peng, et al., “Heterogeneous cloud radio access networks: a new perspective for
enhancing spectral and energy efficiencies,” IEEE Wireless Commun., Dec. 2014
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 15

1. 5G Vision and Architectures
5G: Why SDN, NFV, Cloud technologies in 5G?
Architecture and technology
SDN

NFV

Cloud
computing

Fog/ Edge
computing

x

x

x

x

x

x

x

x

x

x

x

x

5G Challenge/Problem
High capacity

x

Scalability and flexibility

x

x

User centricity
Programability

x

Self-healing infrastructures

x

Heterogeneity of RATs

x

x
x

Interference mitigation

x

Low latency

x

Energy saving

x

x

Flexible management

x

x

x

x

Wide range of supported
services

x

x

x

x

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 16

CONTENTS
1. 5G Vision and Architectures
2.
Software Defined Networking (SDN) and Network
Function Virtualization (NFV)
3. Cloud Computing Architectures in 5G
4. SDN and NFV in 5G
5. Fog Computing in 5G
6. Mobile Edge Computing
7. Open research topics
8. Conclusions

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 17

2. Software Defined Networking and
Network Function Virtualization
2.1 SDN main objectives and features
Recent industry/research effort - results:
SDN –new networking architecture
Open Networking Foundation (ONF- non-profit industry consortium )
OpenFlow I/F specs for SDN

several

Promises for enterprises, data centres, carriers :
higher programmability, automation, and network control
highly scalable, flexible networks
fast adaptation to changing business needs
SDN objectives:
Control Plane (CPl) and Data Plane (DPl) separation
A centralized logical control and view of the network
• underlying network infrastructure is abstracted from the applications
• common APIs
Open I/Fs between the CPl (controllers) and DPl elements.
Network programmability: by external applications including network
management and control
Independency of operators w.r.t. network equipment vendors
Technology to be used in Cloud data centers as well in WANs
Increased network reliability and security
OpenFlow : typical (“vertical”) protocol DPl ---CPl
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 18

2. Software Defined Networking and Network
Function Virtualization
2.2 SDN Basic
Architecture
Mgmt.
Application Plane

Application
Routing

Plane

Network OS:
Distributed system that
creates a consistent,
updated network view
Executed on servers
(controllers) in the
network
Examples: NOX, PoX,
ONIX, HyperFlow,
Floodlight, Trema,
Kandoo, Beacon,
Maestro,..
SDN controller uses
forwarding abstraction in
order to:
Collect state information
from forwarding nodes
Generate commands to
forwarding nodes

SLA
contract

Application
QoS control

Application
Traffic engineering

Abstract
network
view

Northbound
I/F

Network Abstraction/Virtualisation

Control
Plane
Network OS
Config.
Policies
Monitoring
Southbound
I/F

Consistent
updated
global
network view

Open I/F to
Forwarding Plane
e.g. OpenFlow
Swich/
Router

Node
setup

Swich/
Router

Swich/
Router

Swich/
Router
Swich/
Router

Data
Plane

Swich/
Flow Table
Fwd. Engine

Router

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 19

2. Software Defined Networking and Network
Function Virtualization
2.3 Network Function Virtualization
NFV objectives:
Improved capital efficiencies vs. dedicated HW implementation solutions, by:
• Using COTS computing HW to provide Virtualized Network
Functions (VNFs) through SW virtualization techniques
• Sharing of HW and reducing the number of different HW architectures
Improved flexibility in assigning VNFs to HW
better scalability
decouples functionality from location
enables time of day reuse
enhance resilience through Virtualization, and facilitates resource sharing
Rapid service innovation through SW -based service deployment
Common automation and operating procedures ⇒ Improved operational
efficiencies
Reduced power consumption
(migrating workloads and powering down unused HW)
Standardized and open I/Fs: between VNFs infrastructure and mgmt. entities

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 20

2. Software Defined Networking and Network
Function Virtualization (NFV)
2.3 Network Function Virtualization (cont’d)
Network services are provisioned differently w.r.t current networks practice
Decoupling SW from HW
network element is no longer a collection of integrated HW@SW entities
⇒ they may evolve independently
Flexible network function deployment:
The SW/HW detachment allows to reassign and share the infrastructure
resources
HW and SW can perform different functions at various times
The pool of HW resources is already in place and installed at some
NFVI-PoPs ⇒ the actual NF SW instantiation can be automated.
• leverages the different cloud and network technologies currently
available
• helps NOs to faster deploy new network services over the same
physical platform.
Dynamic operation
network function are performed by instantiable SW components ⇒
• greater flexibility to scale the actual VNF performance in a
dynamic way
• finer granularity, for instance, according to the actual traffic
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 21

2. Software Defined Networking and Network
Function Virtualization
2.3 Network Function Virtualization (cont’d)
NFV vision ( source : ETSI)

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 22

2. Software Defined Networking and Network
Function Virtualization
NFV Architecture
High level view of NFV
framework

NFV
Orchestrator

Working domains
VNF, as the SW implementation
of a NF
NFV Infrastructure (NFVI)
includes the PHY resources and
how these can be virtualized
NFVI supports the execution
of the VNFs.
NFV Management and
Orchestration (NFV-MANO)
orchestration and lifecycle
management of physical
and/or SW resources
NFV MANO focuses on all
virtualization-specific
management tasks

Os-Ma

Operation System Support / Bussines System Support
(OSS/BSS)

Virtualised Network Functions
Or-Vnfm

EMn

EM1
Ve-Vnfm

VNF
VNF
Manager
Manager


VNFn

VNF1
Or-Vi
Vi-Vnfm

Service, VNF
Infrastructure
Description

Virtualised
Infrastructure
Manager(s)

Vn-Nf

Vn-Nf
NFVI

Virtual

Computing

Storage

Network

Nf-Vi

Virtualisation Layer
Vi-Ha

NFV Management
and Orchestration

Hardware
Resources

Execution
Reference points

Computing

Main NFV
Reference points

Storage

Network

Other Reference
points

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 23

2. Software Defined Networking and Network
Function Virtualization
2.3 NFV- SDN cooperation
SDN/NFV : complementary technologies
Both build on the rapid evolution of IT and
cloud technologies
SDN features as:
• separation CPl/DPl
• ability to abstract and program
network resources
fit nicely into the NFV paradigm ⇒
• SDN can play a significant role in the
orchestration of the NFV
Infrastructure resources (both
physical and virtual) enabling :
provisioning and configuration of
network connectivity and
bandwidth

automation of operations

security and policy control
The SDN controller maps to the overall
concept of network controller identified in the
NFV architectural framework
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 24

2. Software Defined Networking and Network
Function Virtualization
NFV SDN-Cooperation
ONF: NFV and SDN – industry view on architecture
Source: ONF

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 25

2. Software Defined Networking and Network
Function Virtualization (NFV)
Source: “SDN and
OpenFlow World
Congress”, Frankfurt,
October 15-17, 2013

SDN and Network Function Virtualization
Os-Ma

NFV
Orchestrator

Operation System Support / Bussines System Support
(OSS/BSS)

Virtualised Network Functions
Or-Vnfm

EMn
Ve-Vnfm

VNF
VNF
Manager
Manager

OpenFlow
VNFswitch

OpenFlow
Controller

Or-Vi


VNFn

Vn-Nf
NFVI

Vi-Vnfm
Virtual
Network

Storage

Computing

Virtual

Nf-Vi

Virtualised
Infrastructure
Manager(s)

Virtualisation Layer

Openflow
vSwitch

Config. point

Vi-Ha

NFV Management
and Orchestration

Execution
Reference points

Vn-Nf

Openflow
pSwitch

Computing

Main NFV
Reference points

Storage

Hardware
Resources

Other Reference
points

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 26

CONTENTS
1. 5G Vision and Architectures
2. Software Defined Networking (SDN) and Network
Function Virtualization (NFV)
3.
Cloud Computing Architectures in 5G
4. SDN and NFV in 5G
5. Fog Computing in 5G
6. Mobile Edge Computing
7. Open research topics
8. Conclusions

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 27

3. Cloud Computing Architectures in 5G
Cellular systems evolution towards 5G (cont’d)
CRAN Cloud Radio Access Networks- solution proposed for 5G
CRAN ( interest from academia and industry)
large number of low-cost Remote Radio Heads (RRHs), randomly
deployed and connected to the base band unit (BBU) pool through
the fronthaul links
Advantages:
• RRHs closer to the users
higher system capacity, lower power
consumption
• the baseband processing centralized at the BBU pool
cooperative
processing techniques to mitigate interferences
• exploiting the resource pooling and statistical multiplexing gain
efficiency in both energy and cost

Drawbacks:

• the fronthaul constraints have great impact on worsening perf. of
CRAN, and the scale size of RRHs
• accessing the same BBU pool is limited and could not be too large due to
the implementation complexity

Note: many architectures are proposed by different mobile operators,
manufactories, researching institutes
an unified CRAN for 5G is
still not straightforward
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 28

3. Cloud Computing Architectures in 5G
Cellular systems evolution towards 5G (cont’d)
H-CRAN Heterogeneous Cloud Radio Access Networks
HetNet
Low Power Nodes (LPN) ( e.g., pico BS, femto BS, small BS , etc.)
are key components to increase capacity in dense areas with high
traffic demands.
High power node (HPN), e.g., macro or micro BS) combined with
LPN to form a HetNet
Problem: too dense LPNs - >interferences,
need to control
interferences
• Method : advanced DSP techniques
• 4G solution: The coordinated multi-point (CoMP)
• (-) in real networks because CoMP performance gain depends
heavily on the backhaul constraints
• Conclusion: cooperative processing capabilities is needed in the
practical evolution of HetNets

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 29

3. Cloud Computing Architectures in 5G
H-CRAN (cont’d)
Notes:
In 1G, 2G, 3G: cooperative processing is not needed  the inter-cell interference
can be avoided by utilizing static frequency planning or CDMA
4G - OFDM-based: intercell interference is severe
cooperative processing through CoMP is critical

intercell or inter-tier

H-CRAN-based 5G system
Cloud computing based cooperative processing and networking
techniques are proposed to tackle the 4G challenges alleviating inter-tier
interference and improving cooperative processing gains
the HPNs are enhanced with massive multiple antenna techniques and
simplify LPNs through connecting them to a “signal processing cloud” with high
speed optical fibers
The baseband datapath processing + LPNs radio resource control are
moved to the cloud server
• cloud computing based cooperation processing and networking gains are
fully exploited
• operating expenses are lowered
• energy consumptions of the wireless infrastructure are decreased

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 30

3. Cloud Computing Architectures in 5G
5G System Architecture in H-CRAN approach
Simplified H-CRAN architecture
Gateway

BBU Pool

Internet

BBU
BBU

RRH – Remote Radio Head;
HPN – High Power Node
LPN- Low Power Node
BBU- baseband (processing)
unit

Backhaul
Fronthaul

HPN

RRH

MT

RRHs include only partial PHY
functions ;
The model with these partial
functionalities is denoted as PHY_RF

RRH

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 31

3. Cloud Computing Architectures in 5G
5G System Architecture in H-CRAN approach

RRH – Remote Radio Head;
X2/S1 – 3G imported interfaces
HPN – High Power Node
LPN- Low Power Node
BBU- baseband (processing)
unit
BSC- Base Station Controller
(2G/3G)
MIMO – Multiple Inputs –
Multiple Outputs
LTE – Long Term Evolution ( 4G)

Source: M. Peng, et al., “Heterogeneous cloud radio access networks: a new perspective for
enhancing spectral and energy efficiencies,” IEEE Wireless Commun., Dec. 2014
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 32

3. Cloud Computing Architectures in 5G
Cellular systems evolution towards 5G (cont’d)
5G HetNet Solution (details)
Increase the capacity of cellular networks in dense areas with high
traffic demands
Key components in HetNets: Low Power Nodes (LPN) which serve
for the pure “data-only” service with high capacity
Advantages:
HetNets decouples the control plane and user plane
LPNs only have a very simple control plane, while the control channel
overhead and cell-specific reference signals of LPNs can be fully
shifted to Macro Base Stations (MBSs)
Drawbacks:
an underlaid structure that MBSs and LPNs reuse the same spectral
resources
severe inter-tier interferences
• it is critical to suppress interferences through advanced DSP
by adopting the advanced Coordinated Multi-point (CoMP)
transmission and reception technique to suppress both intra-tier and
inter-tier interferences.

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 33

3. Cloud Computing Architectures in 5G
H-CRAN ( cont’d)
H-CRAN-based 5G system (details)

The RRHs : relay (by compressing and forwarding) the received
signals from UEs to the centralized baseband unit (BBU) pool
through the wired/wireless fronthaul links
• There is a high number of RRH with low energy consumption
• Perform only the front RF and simple symbol processing
• Other important baseband PHY processing and procedures of
the upper layers are executed jointly in the BBU pool
The joint decompression and decoding are executed in the BBU
pool
HPNs are still critical in C-RANs to
• guarantee backward compatibility with the existing cellular
systems
• support seamless coverage since RRHs are mainly deployed to
provide high capacity in special zones
The HPNs, help the convergence of multiple heterogeneous radio
networks
• all system control signaling is delivered wherein.
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 34

3. Cloud Computing Architectures in 5G
5G System Architecture in H-CRAN approach (cont’d)
The I/Fs :[ BBU pool – HPNs]
mitigate the cross-tier interference
RRHs - HPNs through centralized CC-based cooperative processing
techniques.
The data and control I/F (BBU pool – HPNs) are S1 and X2, respectively
H-CRAN supported services: voice and data
voice service admin - HPNs
high data packet traffic is mainly served by RRHs.
Participation of HPNs
H-CRAN alleviates the front-haul requirements
The control signaling and data symbols are decoupled in H-CRANs.
Favours a SDN-like approach

All control signaling and system broadcasting data are delivered by
HPNs to UEs
which simplifies the capacity and time delay constraints in the (BBU
pool – RRHs) fronthaul links
and makes RRHs active or sleep efficiently to decrease energy
consumption
burst traffic or instant messaging service with a small amount of data
can be supported efficiently by HPNs
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 35

3. Cloud Computing Architectures in 5G
5G System Components in H-CRAN approach
Cloud computing technologies
on_ demand resource processing,
storage, and network capacity
wherever needed
Software-defined air interfaces and
networking technologies are
integrated
the flexibility to create
new services and applications
RRH – Remote Radio Head;
ACE - Ancestral Communication Entity i.e. :
MBSs, micro BSs, pico BSs, etc.)
HPN – High Power Node
MIMO – Multiple Inputs –Multiple Outputs

Source:M. Peng et al., “System Architecture and Key Technologies for 5G Heterogeneous

Cloud Radio Access Networks,” IEEE, Network, vol. 29, no. 2, Mar. 2015, pp. 6–14.
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 36

3. Cloud Computing Architectures in 5G
5G System Components in H-CRAN approach

H-CRANs uses CC + heterogeneous convergence technologies
New entity Node C (Node with CC )
~ to 3GPP BS evolution
has to converge different RANs for comm. entities (ACEs, i.e. MBSs, micro BSs, pico
BSs, etc.)
processing and network functionalities in the PHY and upper layers for the newly
designed RRHs
1. Node C works to converge ACEs, it is ~ convergence GW, to execute:
the cooperative multiple-radio resource managements (CM-RRM)
media independent handover (MIH) functionalities
Can play role of traditional (RNC) and BS controller (BSC)
2. Node C is used to manage RRHs: it acts as the BBU pool, which is inherited from
CRANs.
Node C has powerful computing capabilities to execute large scale cooperative:
signal processing in the PHY
networking in the upper layers
RRHs mainly provide high speed data transmission ; no CPl in hot spots
The control channel overhead and cell specific reference signals for the whole HCRAN are delivered by ACEs.
UEs nearer to ACEs than RRHs are served by ACEs and called HUEs
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 37

3. Cloud Computing Architectures in 5G
5G System Components in H-CRAN approach
5G H-CRAN = UEs, H-CRAN, and IoTs (details)
Three architectural Planes:
User/Data Plane (U) carries the actual user traffic, related traffic processing
Control Plane (C) - control sgn. and resource allocation and traffic processing
to improve SE and EE.
Management Plane (M)
• administration and operation,
• add/delete/update/modify the logic and interactions for the U and the C
The H-CRAN architecture is software defined; it has attributes of SDN and CC
overall system components – heterogeneous set:
User Equipments, IoT Devices
Network infrastructure – different technologies (MBS, microBS, picoBS, Access Points,
Routers, etc.
Node C can play also the SDN controller role

Applications (on top of SDN logical infrastructure)
Management plane:
• Self-organizing : Minimum drive test, Inter and Intra network SON
• Resource cloudification: Cell association, user-centric scheduling, power control,
load/handover control
Control plane: Cognitive processing: Underlaid, overlaid, hybrid
User Plane: Big data mining, Machine learning, traffic-driven and user-centric optimization
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 38

CONTENTS
1. 5G Vision and Architectures
2. Software Defined Networking (SDN) and Network
Function Virtualization (NFV)
3. Cloud Computing Architectures in 5G
4.
SDN and NFV in 5G
5. Fog Computing in 5G
6. Mobile Edge Computing
7. Open research topics
8. Conclusions

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 39

4. SDN and NFV in 5G
Example 1: 5G network generic architecture [*]
Operator high level services, OTT

API
NFV based network cloud
Network Intelligence

Internet
Data (User)
Plane

Control
Plane

Data Path backhaul
(fiber, copper, cable)
Control Path backhaul

BS

RRU

RRU

Data Path
(wireless)

Macro Cell
Control Path
(wireless)

Small
Cell

D2D
MT

Train

RRU –Remote Radio Unit; D2D – Network controlled Device to Device;
MTC – Machine type Communication; OTT- Over the Top; MT – Mobile Terminal;
NFV- Network Function Virtualisation; API- Application Programmer Interface
[*] See Ref. : Agyapong P.K., Iwamura M., Staehle D., Kiess W., Benjebbour A. ’
Design Considerations for a 5G Network Architecture’. IEEE Communications Magazine, November 2014, pp. 65-75
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 40

4. SDN and NFV in 5G
Example 1: 5G network generic architecture – details
Two logical network layers
Network cloud - higher layer functionalities (different functions could
be dynamically instantiated and scaled based on SDN/NFV)
Radio Network (RN) - a minimum set of lower layers L1/L2
functionalities
Three main design concepts integrated:
NFV and SDN with control/user plane split, to provide flexible
deployment and management/operation;
ultra-dense small cell deployments (licensed/unlicensed spectrum), to
support high capacity and data rate challenges
the network data are used in the cloud, to optimise resources usage
and for QoS provisioning and planning.

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 41

4. SDN and NFV in 5G
Example 1: 5G network generic architecture – details (cont’d)
Characteristics/design solutions
A redesigned stack integrates Access Stratum (AS) and Non Access
Stratum (NAS).
Splitting the Control/User (data) planes and using different frequency
bands for coverage and capacity.
Relaying and nesting configuration - in order to support multiple
devices, group mobility, and nomadic hotspots.
The network intelligence is data-driven, to optimise of the network
resource planning and usage.
Connectionless and contention-based access is proposed with new
waveforms for asynchronous access of massive numbers of
machine-type communications (MTC)
• connected cars, connected homes, moving robots, and sensors

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 42

4. SDN and NFV in 5G
Example 1: 5G network generic architecture – details (cont’d)
Characteristics/design solutions (cont’d)
The NFV based network cloud is split into CPl and DPl (following the
SDN principle) and a ‘network intelligence- NI’ layer could be put on top
of them.
CPl : mobility management, radio resource control, NAS-AS
integration and security functions (e.g. authentication, etc.)
DPl (User Plane) assures the data flow paths between different
RANs and to/from Internet.
• e.g. gateway functions, data processing functions, mobility
anchors, security control on the air interface, etc.
NI: services orchestration ( traffic optimisation, QoS provisioning,
caching control, etc.)
• It can analyse the big data collected from the different
components (core, RAN) and infer appropriate actions
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 43

4. SDN and NFV in 5G
Example 1: 5G network generic architecture – details (cont’d)
Characteristics/design solutions (cont’d)
CPl and DPl instances can be seen as “data centers” having high amount
of resources.
Each data centre can control one or several macrocells and/or RRUs
The DPl and CPL entities could be located close to BSs and also to
RRUs, if some latency- critical services requirements should be met
The operator can deploy both large and small data centres to support
specific service needs
BSs are more simple and more energy-efficient w.r.t. conventional 4G
case

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 44

4. SDN and NFV in 5G
Example 1: 5G network generic architecture – details (cont’d)
Characteristics/solutions (cont’d)
The network cloud allows for resource pooling, reducing overprovisioning and under-utilisation of network resources.
SDN + NFV -> dynamic deployment and scaling on demand of network
functions.
the local data centres can borrow resources from each other (when traffic
load is low)
they also can be enriched (installing new software) to support other apps
the cloud-computing model flexibility is present in the network cloud

Open issue: decision on balancing between different allocation of
functions and specifically how to incorporate small cells with NFV and
SDN in a cost effective manner use in small cells in different frequency
regimes

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 45

4. SDN and NFV in 5G
Example 2: Cellular SDN architecture (CSDN) [*]
CSDN is using the SDN and NFV principles
Goal : to optimise the dynamic resource orchestration, by performing real-time
context data gathering, analysis and then making intelligent decisions.
CSDN :forwarding, control, and network application architectural planes
The network and user information are collected from the mobile edge networks
used locally, or exported/ shared to other service providers, to enrich the set of
services
A novel Knowledge Plane is added
to cooperate with network application plane
Mobile Services Provider (MSP) can construct an intelligent vision upon its
network and users’ environment
New apps. or virtual functions can be implemented and instantiated
(e.g., optimised content distribution and caching, Internet of Things (IoT),
location based services, etc.)
and linked to the controller northbound interface
[*] Source ref.: Bradai A., Singh K., Ahmed T., and Rasheed T., ‘Cellular Software Defined Networking: A framework’.
EEE Communications Magazine — Communications Standards Supplement, June 2015, pp. 36-43
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 46

4. SDN and NFV in 5G
Example 2: Cellular SDN architecture (CSDN) ([*])
Knowledge Plane
Application
Plane

VPCRF

VMME

Other
Applications

VEPC
VeNB
VP-GW

VS-GW

LTE functional blocks:
eNB
Evolved Node (Base
Station)
MME
Mobility Management Entity
S-GW
Serving gateway
PCRF
Policy Control and Charging
Rules Function
P-GW
Packet Data Network(PDN)
Gateway

Northbound
I/F

Network Abstraction/Virtualisation
Topology
Discovery

Topology resource
view

Network
Monitoring

Control
Plane

Other
Controllers

Network OS

Southbound
I/F

Data Plane
Switch
Switch

Switch
Switch

Other
packet
networks

[*] Source ref.: Bradai A., Singh K., Ahmed T., and Rasheed T., ‘Cellular Software Defined Networking: A framework’.
EEE Communications Magazine — Communications Standards Supplement, June 2015, pp. 36-43
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 47

4. SDN and NFV in 5G
Example 2: Cellular SDN architecture (CSDN) – details
The CSDN example is oriented towards the 4G LTE:
several functions can be implemented as VNFs at the CSDN application level,
in a centralised cloud-based infrastructure

The sub-systems included in the architecture :
LTE Evolved Packet Core (EPC) and eNodeB (eNB)
The LTE virtualised network functionalities interact at the M&C level with
the CSDN switches via the controller
Data plane :
CSDN switches of the ePC and its boundary is placed at the eNB
switches corresponding respectively to eNBs, Serving Gateway (S-GW) and
Packet Data Network Gateway (P-GW)

Main Control Plane component: Network Operating System (NOS) and
an abstraction/virtualisation layer
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 48

4. SDN and NFV in 5G
Example 2: Cellular SDN architecture (CSDN) – details
Control Plane (CPl) SDN interfaces:
north I/F - to the Application Plane, south I/F to the DPl and east-vest I/F
towards other controllers.
Virtual network functions (VNF) are defined in the Application Plane, to execute the
functions of UTRAN and EPC
VNFs, named VeNB, Virtual Mobility Management Entity (VMME) VS-GW, VPGW and Virtual Policy Control and Charging Rules Function (VPCRF),
+ their corresponding switches in DPl (e.g. VeNB plus its CSDN switch
correspond to eNB functionalities), perform the equivalent of LTE UTRAN and
EPC functionalities
Other applications could be added to the Application Plane
Open issues:
wireless network, high number of users, scarce radio resources → fast real
time response is required
centralising all functions in the SDN controller is not scalable
further study for functions allocation balancing between a switch and a controller

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 49

4. SDN and NFV in 5G
Example 2: Cellular SDN architecture (CSDN) – details
Open issues ( cont’d)
How many controllers? One/several controller(s) per which network zone?
Large networks -> need several controllers
How to distribute them geographically?
The controller placement multi-criteria problem is a NP-hard one…
• maximize the controller-forwarder or inter-controller communication
throughput
• limit the controller overload (load imbalance)
• fast recovery after failures (controllers, links, nodes).
Data Plane: Transforming the LTE specific functionalities in packet flow rules
specific to SDN
Real time response of the VMME ?

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 50

4. SDN and NFV in 5G
Example 3: 5G unified Control Plane and Data Plane [*]
5G architecture based on SDN, NFV and edge computing
Three control levels: Device, Edge and Orchestration Controllers,
fully decoupled from the DPl and implementing a unified security, connection,
mobility and routing management for 5G networks
The solution is backward compatible to 3GPP releases
SDN-based connectivity between VNFs (applications) enables carrier grade
communication paths, by avoiding tunnelling
low E2E latency
appropriate for mission critical communications
flexible, reliable and dependable
Implementation variants: ‘centralised’ or ‘distributed at the edge’, depending on
functional and non functional requirements of the supported services
CPl and DPl logical network elements are decomposed into sets of applications or
modules
modules can be dynamically instantiated in the cloud infrastructure according to
network operation or service requirements.
[*] Guerzoni R., Trivisonno R., Soldani D. ‘SDN-Based Architecture and Procedures
for 5G Networks’. 5GU 2014, November 26-27, Levi, Finland, DOI 10.4108/icst.5gu.2014.258052
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 51

4. SDN and NFV in 5G
Example 3: The 5G unified Control Plane and Data Plane [*]
Orchestration
Controller
Edge
Controller
(I)

Device
Controller
(EC II)
Internet

AN
User
Equipment

LHRE
Cloud Infrastructure
AN
AN

AN – Access Node
Control
Plane AS

Control
Plane NAS

MPl

Data
Plane

Control
Plane SDN

AS – Access Stratum; NAS Non-Access Stratum; MPl – Management Plane;
LHRE- Last Hop Routing Element
[*] Guerzoni R., Trivisonno R., Soldani D. ‘SDN-Based Architecture and Procedures
for 5G Networks’. 5GU 2014, November 26-27, Levi, Finland, DOI 10.4108/icst.5gu.2014.258052
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 52

4. SDN and NFV in 5G
Example 3: The 5G unified Control Plane and Data Plane – details
The Device Controller (DC) (located in the device) controls the PHY layer connectivity to
the 5G network and handles AS functions such as access/network selection.
Two types of EDGE Controllers, (EC, (i) and (ii)), implementing
Network Access Control; Packet Routing and Transfer
Radio Resource Management
Mobility and Connection Management ; Security
The EC has similar functions to the AS/NAS 4G functions in eNodeB and MME
The EC implementation is distributed over the cloud infrastructure, being composed of
several interconnected Control Applications (C-Apps)
Each C-App performs a subset of functions, like:
Radio Access (RA), Authorization & Authentication (AA)
Admission Control (AC), Flow Management (FM), Mobility Management (MM),
Connection (Session) Management (CM), Security (Sec).

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 53

4. SDN and NFV in 5G
Example 3: The 5G unified Control Plane and Data Plane – details (cont’d)
To fully separate the DPl/CPl also on the radio link, the RA App is split respectively into
RAD /RAC applications.
The DPl could be instantiated on a different Point of Presence (PoP)
For some mission critical communications, the mobile devices might be required to
support some AS/NAS functions; that is why two types, i.e.,
EC (i) – with C-Apps instantiated in the edge cloud infrastructures and
EC(ii) - implemented temporarily or permanently on a mobile device.
The Orchestration Controller (OC) modules :
Resource Orchestration (RO)
Topology Management (TM) - has network mgmt. functions (similar to those of 4G)
OC coordinates the utilisation of cloud resources (computational, storage and
networking), allocating and maintaining the resources, to instantiate CPl and DPl
RO allocates PHY resources to instantiate EC Control Apps. i.e., it determines
the embedding solution for the virtual CPl and DPl to be instantiated

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 54

4. SDN and NFV in 5G
Example 3: The 5G unified Control Plane and Data Plane – details (cont’d)
Data Plane (DPl):
SDN clean-slate architecture has been adopted
It did not define neither dedicated DPl network elements (e.g., 4G SGW and PGW),
nor unique logical elements (e.g. mobility anchor points)
Advantages of the architecture:
reconfigurability - one can dynamically instantiate logical architectures, implementing
network functions, services and corresponding states in the optimal location within
the cloud infrastructure
No more need of tunnelling protocols (common in 3GPP) (lower fwd. latency)
latency of forwarding paths could be reduced to almost zero by pro-actively
configuring the SDN-based infrastructure (thus realising the “always-on” concept
already present in 4G EPC
The proposed architecture, functions and procedures have the potential to
become an important candidate solution for 5G

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 55

4. SDN and NFV in 5G
Example 4: SoftAir architecture ([*])
SoftAir: a novel software-defined 5G architecture targeting NFV and
cloudification and aiming to scalability, flexibility and resilience
Main characteristics: fine-grained BS decomposition, OpenFlow interfaces,
mobility-aware control traffic balancing, resource-efficient network
virtualisation, and traffic classification
SoftAir architecture : SDN Control Plane and Data Plane
DPl contains a complete network infrastructure: Software Defined Core
Network (SD-CN) and SD – Radio Access Network (SD-RAN)
The OpenFlow and SNMP protocols link the two planes
SD-CN is composed by SD-switches, under CPl coordination
[*] Akyildiz I.F., Wang P., Lin S.C. ‘SoftAir: A software defined networking architecture for 5G
wireless systems’, Computer Networks 85 (2015) pp.1–18.
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 56

4. SDN and NFV in 5G
Example 4: Overall architecture of the SoftAir ([*])
Customised
Application

SDN
Control
Plane

Management
Applications

Network Controller
Open I/F
OpenFlow, SNMP

SDN Data Plane

Core
Network

SD Baseband Servers (BBS)

BU1
SDRAN1

..

BUn
Backhaul
link

Fronthaul
Network

BBS
BU1

RRHs

..

BUn

SD-RAN2

Akyildiz I.F., Wang P., Lin S.C. ‘SoftAir: A software defined networking architecture for 5G
wireless systems’, Computer Networks 85 (2015) pp.1–18.
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 57

4. SDN and NFV in 5G
Example 4: SoftAir architecture (cont’d)
SoftAir supports:
development of customised SDN applications, e.g., mobility management, QoSbased routing, billing policies, etc., in CPl
global management tools and network virtualization.
Current field deployment have shown the SDN advantages (B4 - Google, SWAN Microsoft, ADMCF - Huawei, etc.)
SD-CN can obtained important increase in link utilisation from 30–40 per cent to
over 70 per cent
Solutions for scalability (controller-to–[SD-CN] forwarders) communication
use of high performances controllers and/or
use multi-controller clusters and multi-threading technologies.
recent research : in large – scale SDN networks with in-band control channels, the
controller-forwarder comm. delay can be minimised by using traffic balancing
schemes, based on parallel optimisation theories.

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 58

4. SDN and NFV in 5G
Example 4: SoftAir architecture (cont’d)
SoftAir : mobility-aware and proactive control traffic balancing scheme, minimising the
CPl-DPl delay by exploiting the SD-RAN mobile feature
the control traffic in SD-RAN is following some spatial and temporal patterns
SoftAir supports:
L1-L3 function virtualisation, performed in a distributed architecture
The SD-BS is split into
hardware-only Radio Heads (RHs)
and software-implemented baseband units (these two components could be also
remotely located).
A fronthaul network (fiber/microwave) connects the Remote Radio Heads (RRH) to
Baseband processing Servers (BBS) using standardised interfaces like
Open Base Station Architecture Initiative (OBSAI)
or Common Public Radio Interface (CPRI)
Current standardisation effort Std.entities + industry to make the RRH-BBS technology
independent.
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 59

4. SDN and NFV in 5G
Example 4: SoftAir architecture (cont’d)
SoftAir SD-RAN has some similarities to CRAN Architecture
CRAN
focused on high-performance computing of baseband processing
functions (mostly for L1 operations) at remote servers or data centers
(-) : CRAN cannot achieve scalable PHY/MAC-layer cloudification and
does not support network-layer cloudification as SD-CN does
SD-RAN
offers scalability, evolvability, and cooperativeness through fine-grained
BS decomposition that overcomes fronthaul traffic burden
In SD-RANs partial baseband processing is done at the RRH (e.g.,
modem) while the remaining baseband functions (e.g., MIMO coding,
source coding and MAC), are executed at the BBS

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 60

4. SDN and NFV in 5G
Example 4: SoftAir architecture (cont’d)
SoftAir functional split
SD-RAN (has reduced data rate requirements) is scalable
It also can support cooperative gain
Is evolvable by allowing the aggregation of a large number of technology-evolving
RRHs at BBS and CPRI-supported fronthaul solutions
BSs can be managed in SDN style via a unified interface, for different wireless
standards (multi-technology capability)
Seamless vertical mobility is possible
SoftAir network virtualisation
multiple virtual networks (VNet) on the same PHY infrastructure
each VNet (slice) may independently adopt its L1/L2/L3 protocols
In SoftAir advanced solutions can be deployed on demand and dynamically
allocated

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 61

4. SDN and NFV in 5G
Example 3: Function cloudification in SoftAir system

Cellular Applications:

Management Tools:

Routing,
Mobility management,
Subscriber Data Base

Global traffic learning,
Network Hypervisor,
Control Traffic Balancing

Control
Plane

SDN Network Controller
OpenFlow I/F

Data Plane

SD-BS

control

Traffic
classifier

Forwarding
Table
MAC

PHY ( partial)

SD switch

SD switch
Forwarding
Table

Forwarding
Table

Switching
Fabric

Switching
Fabric

CPRI

SD-CN

data
Modem

Modem

Radio
circuits

Radio
circuits

Antenna

Hardware

Antenna

Software
Open I/F

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 62

4. SDN and NFV in 5G
Example 4: SoftAir architecture (cont’d)
The SoftAir NV enables a wide range of applications.
Each Mobile Virtual Network Operator (MVNO) may use different wireless
technologies (WiFi, WiMAX, LTE, small-cells, HetNets, etc.)
The virtual slices can be customised for different services and types of traffic flows
- e.g., for QoS routing, E2E controlled performances, etc.
The slices isolation might accelerate the innovation, (in a slice, one can
independently develop new protocols)
SoftAir : three types of hypervisors (to realise virtualization):
high-level- network hypervisor
low-level
wireless hypervisors
switch hypervisors
Thus SoftAir enables the E2E network virtualisation traversing both SD-RAN and SDCN, realising a truly multi-service converged network infrastructure.

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 63

4. SDN and NFV in 5G
Example 4: SoftAir architecture (cont’d)
SoftAir architecture advantages
RAN sharing - > CAPEX reduction
Evolutionary adaptive arch. , due to DPl/CPl separation and DPl programmability
SDN: DPl/CPl separation allows HW/SW infrastructures to evolve independently
For instance, novel RATs (e.g., mm-waves, full-duplex, massive MIMO, THz) can
be adopted in hardware
CPl:Traffic engineering and network mgmt. optimisation solutions can be applied
The cloud style and network virtualization creates possibility to offer Infrastructure-asa-Service (IaaS) on top of the same physical network
useful for emerging different network services, e.g., M2M, smart grid, MVNOs,
OTT content services like Netflix video streaming, etc.

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 64

4. SDN and NFV in 5G
Example 4: SoftAir architecture (cont’d)
SoftAir architecture advantages
Distinct Service Providers (SP) can independently control, optimize, and
customize the underlying infrastructure without owning it and without
interfering with other SPs.
Network resources (e.g., spectrum), can be dynamically shared among
SPs, e.g., Mobile Virtual Network Operators (MVNO)s.
A good spectral efficiency can be achieved, due to cooperativeness
Convergence of het-nets networks due to its open and technologyindependent interfaces,
smooth transition among different RATs:
WiFi, WiMAX, LTE, LTE-A, etc.

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 65

CONTENTS
1. 5G Vision and Architectures
2. Software Defined Networking (SDN) and Network
Function Virtualization (NFV)
3. Cloud Computing Architectures in 5G
4. SDN and NFV in 5G
5.
Fog Computing in 5G
6. Mobile Edge Computing
7. Open research topics
8. Conclusions

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 66

5. Fog Computing in 5G
Fog/Edge (FC) computing characteristics [*]:
Fog computing nodes (FCN) are typically located away from the main
cloud data centres, at the edge.
Cloud computing on fog nodes enables low and predictable latency
FCNs
are wide-spread and geographically available in large numbers
provide applications with awareness of device geographical location
and device context.
can cope with mobility of devices
• i.e. if a device moves far away from the current servicing FCN, the fog
node can redirect the application on the mobile device to associate with a
new application instance on a fog node that is now closer to the device.

offer special services that may only be required in the IoT context
(e.g. translation between IP to non-IP transport)
Fog application code runs on FCNs as part of a distributed cloud
application
[*] Fog Computing and Mobile Edge Cloud Gain Momentum Open Fog Consortium, ETSI MEC and
Cloudlets , Version 1.1 Guenter I. Klas Nov 22, 2015
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 67

5. Fog Computing in 5G
Fog/Edge (FC) computing enabled applications [*]
Data plane :
Pooling of clients idle computing/storage/bandwidth resources and local
content
Content caching at the edge and bandwidth management at home
Client-driven distributed beam-forming
Client-to-client direct communications (e.g., FlashLinQ, LTE Direct, WiFi, Direct,
Air Drop)
Cloudlets (mobility-enhanced small-scale cloud data center located at the edge of
the Internet) and micro data-centers
Control plane
Over the Top (OTT) content management
Fog-RAN: Fog driven radio access network
Client-based HetNets control
Client-controlled Cloud storage
Session management and signaling load at the edge
Crowd-sensing inference of network states
Edge analytics and real-time stream-mining
[*] Source: M.Chiang, "Fog Networking: An Overview on Research Opportunities"
December 2015, https://arxiv.org/ftp/arxiv/papers/1601/1601.00835.pdf
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 68

5. Fog Computing in 5G
FC provides
light-weight cloud-like facility close of mobile users
users with a direct short-fat connection versus long-thin mobile cloud connection
customized and engaged location-aware services

FC is still new and there is lack of a standardized definition
Comparison between Fog/Edge (FC) and Conventional Cloud Computing [*]:

[*] T H. Luan et.; al. , "Fog Computing: Focusing on Mobile Users at the Edge"
arXiv:1502.01815v3 [cs.NI] 30 Mar 2016
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 69

5. Fog Computing in 5G
C-RAN limitation in 5G context
strong fronthaul network requirements (to access the centralised (BBU) pool
high bandwidth and low latency inter-connection fronthaul is necessary
(expensive - in practice)
H-CRAN limitation in 5G context
(+) H-CRAN solves some C-RAN problems
user /data (DPl) and control planes (CPl) are decoupled
in H-CRANs the centralized control function is shifted from the BBU pool (like
in C-RANs) to the high power nodes HPN
(HPNs) are mainly used to provide seamless coverage and CPl functions
RRHs provide high speed data rate for DPl
HPNs are connected to the BBU pool via the backhaul links for interference
coordination
(-) H-CRAN still has some challenges in practice
popular location-based social applications
data traffic peaks over the
fronthaul (RRHs BBU pool)
high transient load for the fronthaul
deploying of a high number of fixed RRHs and HPNs in H-CRANs to meet
traffic peak requirements is not efficient (traffic is low for long time intervals)
do not take full advantage of processing and storage capabilities in edge
devices ( e.g., RRHs and “smart” user equipments (UEs) )
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 70

5. Fog Computing in 5G
Fog/Edge Computing in 5G context
extends the traditional cloud computing paradigm to the network edge
a high amount of storage, comm., control, configuration, measurement and
management is performed at the network edge
collaboration radio signal processing (CRSP) can not only be executed in H-CRANs
centralized BBU pool, but also can be hosted at RRHs and even “smart” UEs
UEs might download packets from closer points ( UEs or RRHs)
integration possibility of on-device processing and cooperative radio resource
management (CRRM) on new types of “smart” UEs
Fog computing based RAN (F-RAN) architecture
real-time CRSP and flexible CRRM at the edge devices
F-RANs can be adaptive to the dynamic traffic and radio environment
lower burden on the fronthaul and BBU pool
achievable user-centric objectives : adaptive technique among (D2D), wireless
relay, distributed coordination, and large-scale centralized cooperation

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 71

5. Fog Computing in 5G
Fog/Edge Computing in 5G context
Fog computing based RAN (F-RAN) architecture
mobile FC and edge cloud can offer new services for information sciences and
Internet of Things (IoT)
design of mobile fog as a programming model for large-scale, latency sensitive
applications in the IoT
FC in 5G environments- open research issue
F-RAN architectures
three layers: cloud computing, network access and terminal layer
F-RAN takes full advantages of the convergence of cloud computing,
heterogeneous networking, and fog computing
The FC network is actually composed of
F-APs (residing in the network access layer) and
F-UEs (placed in the terminal layer)
• F-UEs can inter-communicate (direct D2D mode or through additional FUEs playing the role of mobile relays)
The network access layer is composed by F-APs, HPNs and RRHs

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 72

5. Fog Computing in 5G
F-RAN simplified architecture

BBU – BaseBand Unit
F-AP Fog Access Point
RRH – Remote Radio Head
HPN – High Power Node

Cloud Computing
Network Layer
(Centralized storage and
comm. cloud)

BBU Pool
BBU1

BBUn

..

Backhaul

Backhaul

Fronthaul

Core
Network

( Base Station)

F-UE Fog capable user
equipment

Access Layer

RRHs

F-APs

HPN

(Distributed storage cloud)

Terminal Layer
(Distributed community
cloud)

UEs

F-UEs

UEs

Logic Fog Layer

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 73

5. Fog Computing in 5G
F-RAN architecture- example

BBU – BaseBand Unit
F-AP Fog Access Point
RRH – Remote Radio Head
HPN – High Power Node
( Base Station)

F-UE Fog capable user
equipment

Source: M.Peng,S.Yan, K.Zhang, and C.Wang, "Fog Computing based Radio Access Networks:Issues and
Challenges", IEEE NETWORK,2015, http://arxiv.org/abs/1506.04233

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 74

5. Fog Computing in 5G
F-RAN architecture- example (cont’d)

Source: M.Peng,S.Yan, K.Zhang, and C.Wang, "Fog Computing based Radio Access Networks:Issues and
Challenges", IEEE NETWORK,2015, http://arxiv.org/abs/1506.04233

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 75

5. Fog Computing in 5G
F-RAN Architecture features
F-APs
used in the Data Plane to forward and process the traffic data
communicate with BBU pool through the fronthaul links and HPN through
backhaul links

The signals over fronthaul links are large-scale processed in the BBU pool,
while over the backhaul links only control information is exchanged between
the BBU pool and HPN
The BBU pool plays a similar role as in H-CRANs (can also make
centralised caching)
F-RAN alleviates the tasks of the BBU pool and fronthaul links, given that a
large number of CRSP and CRRM functions are shifted towards F-APs and
F-UEs
F-APs and F-UEs may perform limited caching

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 76

5. Fog Computing in 5G
Architectures comparison: C-RAN, H-CRAN and F-RAN
BBU pool and fronthaul burden:
C-RAN –highest; H-CRAN – medium; F-RAN- lowest
F-RAN : Lowest latency
Decoupling between the CPl and DPl : only in H-CRAN and F-RAN
Caching and CRSP functions
centralised in CRAN and H-CRAN
F-RAN- it can be mixed, i.e., centralised/distributed
CRRM functions
centralised in CRAN
H-CRAN F-RAN : mixed solution can be used
Complexity
CRAN or H-CRAN put high complexity in BBU pool and low complexity in
RRHs and UEs
F-RAN exposes medium complexity in BBU pool, F-APs and F-UEs.
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 77

CONTENTS
1. 5G Vision and Architectures
2. Software Defined Networking (SDN) and Network
Function Virtualization (NFV)
3. Cloud Computing Architectures in 5G
4. SDN and NFV in 5G
5. Fog Computing in 5G
6.
Mobile Edge Computing
7. Open research topics
8. Conclusions

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 78

6. Mobile Edge Computing
Why MEC?
MEC provides IT and cloud-computing capabilities within the RAN in
close proximity to mobile subscribers
MEC accelerates content, services and applications so increasing
responsiveness from the edge
Main standardization actors: ETSI, 3GPP, ITU-T
RAN edge offers a service environment with ultra-low latency and highbandwidth as well as direct access to real-time radio network information
(subscriber location, cell load, etc.) useful for applications and
services to offer context-related services
Operators can open the radio network edge to third-party partners
Proximity, context, agility and speed can create value and opportunities
for mobile operators, service and content providers, Over the Top (OTT)
players and Independent Software Vendors (ISVs)
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 79

6. Mobile Edge Computing
MEC Use Cases examples ( content- oriented)
RAN-aware Content Optimization
The application exposes accurate cell and subscriber radio interface information
(cell load, link quality) to the content optimizer, enabling dynamic content
optimization, improving QoE, network efficiency and enabling new service and
revenue opportunities.
Dynamic content optimization enhances video delivery through reduced stalling,
reduced time-to-start and ‘best’ video quality.

Source: https://portal.etsi.org/Portals/0/TBpages/MEC/Docs/Mobileedge_Computing_-_Introductory_Technical_White_Paper_V1%2018-09-14.pdf
Mobile-Edge Computing – Introductory Technical White Paper
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 80

6. Mobile Edge Computing
MEC Use Cases examples ( content- oriented) (cont’d)
Video Analytics
distributed video analytics solution: efficient and scalable mobile solution for LTE
The video mgmt. application transcodes and stores captured video streams from
cameras, received on the LTE uplink
The video analytics application processes the video data to detect and notify
specific configurable events e.g. object movement, lost child, abandoned
luggage, etc.
The application sends low bandwidth video metadata to the central operations
and management server for database searches. Applications : safety, public
security to smart cities

Same source as
in previous slide

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 81

6. Mobile Edge Computing
MEC Use Cases examples ( content- oriented) (cont’d)
Distributed Content and DNS Caching
A distributed caching technology can provide backhaul and transport savings and
improved QoE.
Content caching could reduce backhaul capacity requirements by ~35%
Local DNS caching can reduce web page download time by ~20%

Same source as
previous slide
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 82

6. Mobile Edge Computing
MEC Use Cases examples (content- oriented)
Augmented Reality (AR) content delivery
An AR application on a smart-phone or tablet - overlays augmented reality content
onto objects viewed on the device camera
Applications on the MEC server can provide local object tracking and local AR
content caching;
RTT is minimized and throughput is maximized for optimum QoE
Use cases: offer consumer or enterprise propositions, such as tourist information, sporting
event information, advertisements etc.

Source: https://portal.etsi.org/Portals/0/TBpages/MEC/Docs/Mobile-edge_Computing__Introductory_Technical_White_Paper_V1%2018-09-14.pdf
Mobile-Edge Computing – Introductory Technical White Paper
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 83

6. Mobile Edge Computing
MEC Use Cases examples
Application-aware cell performance optimization
Applied for each device in real time can improve network efficiency and customer
experience
It can reduce video stalling and increase browsing throughput.
Reduce latency
Provide independent metrics on application performance (video stalls, browsing
throughput, and latency) for enhanced network management and reporting

Source: https://portal.etsi.org/Portals/0/TBpages/MEC/Docs/Mobile-edge_Computing__Introductory_Technical_White_Paper_V1%2018-09-14.pdf
Mobile-Edge Computing – Introductory Technical White Paper
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 84

6. Mobile Edge Computing
MEC Use Cases examples
Internet of Things (IoT)
IoT generates additional messaging on telecoms networks, and requires gateways to aggregate
the messages and ensure security and low latency
real time capability is required and a grouping of sensors and devices is needed for efficient
service.
IoT devices are often low in terms of processor and memory capacity
need to aggregate
various IoT messages connected through the mobile network close to the devices
This also provides an analytics processing capability and a low latency response time.

Yun Chao Hu et.al., "Mobile Edge Computing A key technology towards 5G" ETSI White Paper No. 11
September 2015, ISBN No. 979-10-92620-08-5
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 85

5. Mobile Edge Computing
Possible Deployment Scenarios (ETSI)
The MEC server can be deployed in several variants
Note: the multi-technology (LTE/3G) cell aggregation site can be indoor or outdoor

MEC at the LTE macro base station
(eNB) site

MEC at the multi-technology (3G/LTE)
cell aggregation site

MEC at the 3G Radio Network Controller
(RNC) site

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 86

6. Mobile Edge Computing
MEC Architectures
MEC provides a highly distributed computing environment that can be
used to deploy applications and services as well as to store and process
content in close proximity to mobile users.
Applications can benefit from real-time radio and network information and
can offer a personalized and contextualized experience to the mobile
subscriber.
The mobile-broadband experience is more responsive and opens up new
monetization opportunities. This creates an ecosystem where new
services are developed in and around the BS
Key element : (MEC) IT application server which is integrated in RAN
(as above)
The MEC server provides computing resources, storage capacity, connectivity, and
access to user traffic and radio and network information

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 87

5. Mobile Edge Computing
MEC Platform Overview ( source: ETSI)- NFV inspired arch

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 88

CONTENTS
1. 5G Vision and Architectures
2. Software Defined Networking (SDN) and Network
Function Virtualization (NFV)
3. Cloud Computing Architectures in 5G
4. SDN and NFV in 5G
5. Fog Computing in 5G
6. Mobile Edge Computing
7.
Open research topics
8. Conclusions

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 89

7. Open research topics
SDN/NFV
the centralised SDN nature
resilience and scalability

bottlenecks and thus can reduce the

a balance between centralised logical control and actual distributed
infrastructure of controllers should be found
flat or hierarchical architecture of SDN control plane – with multiple
controllers should be adapted to 5G both in Core and RAN
the flat organisation of the SDN controllers does not provide an
effective and flexible management solution for 5G networks to meet
the requirements for resilience and scalability.
different reconfiguration policies should be applied to the network
elements in a dense environment, at different time scales
due to the dynamicity and density of this network
this can also result in high signalling overhead

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 90

7. Open research topics
SDN/NFV (cont’d)
RAN link quality is usually unreliable and unstable, interrupting temporarily
the communication between the controller and its forwarders (if the
controller communication channel uses in-band signalling),
isolated wireless networks problem to solve
5G network might have cells with particular configuration policies, which
should be considered in a differentiated way by the SDN controllers.
the partition of functions to be implemented in each plane is still an open
issue in the SDN/NFV/5G, particularly in the RAN area.
the edge heterogeneity (including D2D, M2M, and V2V) -> very dynamic
topologies -> complexity in SDN and NFV functions planning, increased by
several distinct mobility models and hardware constraints
(e.g., the SDN controller should instruct the switches or network
hypervisor which terminal node should forward packets)
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 91

7. Open research topics
SDN/NFV (cont’d)
integrating SDN and NFV
the SDN programmability needs standardising the N/S interfaces
between physical and virtual network functions that form a single
network service chain.
virtualisation might negatively impact the virtual LTE and Wi-Fi services ->
the VNFs performance should be analysed (to decide about
physical/virtual implementation option.
standardisation of NFV/SDN is still in-progress
a unified cellular programmable interface for implementing SDN and
NFV is under development, including a service chain through the
integration of SDN and NFV

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 92

7. Open research topics
C-RAN/H-CRAN/Fog
two major problems in both CRANs and H-CRANs
high transmission delay and heavy burden on the fronthaul
C-RAN s H-CRANs do not take benefit from processing and storage capabilities
in edge devices, such as RRHs and even ‘smart’ mobile terminals / user
equipments (UEs
use SDN style of control in F-RAN environment?
The combination of the MAC functions and L1 functions for edge devices in FRANs is still not yet clarified
SDN is centralisation-based (for control), while the F-RAN has a distributed
characteristic, based on edge devices
using SDN control for F-RANs-> need to carefully define slices to isolate the
signal processing from resource management in edge devices, as to provide
non-interfering networks to different coordinators.
If SDN controllers are located in cloud computing network layer, -> control traffic
overhead appears (CPl --DPl) to be transported over fronthaul links -> decreasing
the advantages of F-RANs.

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 93

CONTENTS
1. 5G Vision and Architectures
2. Software Defined Networking (SDN) and Network
Function Virtualization (NFV)
3. Cloud Computing Architectures in 5G
4. SDN and NFV in 5G
5. Fog Computing in 5G
6. Mobile Edge Computing
7. Open research topics
8.
Conclusions

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 94

7. Conclusions
5G networking and services – promising technology for a large range of
applications
Cloud, fog, SDN, NFV, MEC, virtualization concepts and technologies can offer
strong support in developing 5G architectures and implementations
Significant effort still to be done to
Find the best architectural configuration in order to realize synergic cooperation
between technologies and take the best benefits from their advantages
Explore further the SDN/NFV cooperation
Distributed & hierarchical architectures for both SDN & NFV
Function split and migration in both SDN and NFV
Explore hierarchical SDN control in 5G+NFV context
Orchestration and management
NFV
Evaluate real time response of the VNFs
Dynamic VNF moving properties in mobility context
MEC – promising technology: it can transform BSs into intelligent service hubs ,
capable of delivering highly personalized services directly from the very edge of the
network while providing the best possible performance in mobile networks.

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 95

Thank you !
Questions?

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 96

List of Acronyms
ACE
AVC
BBU
BS
BSC
BSS
CC
CCN
CDN
CDNP
COTS
CoMP
CP
CRAN
DASH
DRM
EMS
EPC
ETSI
FEC
HCRAN
HPN
HSPA
HTTP
IaaS

Ancestral Communication Entity
Audio Video Conference
Baseband Processing Unit
Base Station
Base Station Controller (2G/3G)
Business Support System
Cloud Computing
Content Centric Neytworking
Content Delivery Network
Content Delivery Network Provider
Commercial-off-the-Shelf
Coordinated multi-point
Content Provider
Cloud RAN
Dynamic adaptive streaming over HTTP
Digital Rights Management
Element Management System
Evolved Packet Core
European Telecommunications Standards Institute
Forward Error Correction
Heterogeneous CRAN
High Power Node
High Speed Packet Access
Hyper Text Transport Protocol
Infrastructure as a Service

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 97

List of Acronyms (cont’d)
IMS
ISG
IT
LPN
LTE
MBMS
M&O
MME
MIMO
NAT
NF
NFV
NFVI
NO
NP
NS
OSS
PaaS
PoC
RAN
RRH

IP Multimedia System
Industry Specification Group.
Information Technology
Low Power Node
Long Term Evolution
Multicast Broadcast Media Service
Management and Orchestration
Mobility Management Entity
Multiple Inputs –Multiple Outputs
Network Address Translation
Network Function
Network Functions Virtualization
Network Functions Virtualization Infrastructure
Network Operator
Network Provider
Network Service
Operations Support System
Platform as a Service
Proof of Concept.
Radio Access Network
Remote Radio Head

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 98

List of Acronyms (cont’d)
RNC
RTP
RTCP
RTSP
SaaS
SDN
SDP
SDO
SLA
S/P-GW
SP
TCP
UDP
VM
VNF

Radio Network Controller
Real Time Protocol
Real Time Control Protocol
Real Time Streaming Protocol
Software as a Service
Software Defined Network
Session Description Protocol
Standards Development Organisation
Service Level Agreement
Serving and Packet Data Networks Gateway
Service Provider
Transmission Control Protocol
User Datagram Protocol
Virtual Machine
Virtual Network Function

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 99

References-1
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.

Andrews J.G., et al., ‘What Will 5G Be?’. IEEE Journal on Selected Areas in Communications, Vol. 32,
No.6, June 2014, pp. 1065- 82
Panwar N., Sharma S., Singh A. K. ‘A Survey on 5G: The Next Generation of Mobile Communication’.
Accepted in Elsevier Physical Communication, 4 Nov 2015, Available from
http://arxiv.org/pdf/1511.01643v1.pdf
Agyapong P.K., Iwamura M., Staehle D., Kiess W., Benjebbour A. ’Design Considerations for a 5G
Network Architecture’. IEEE Communications Magazine, November 2014, pp. 65-75
5G Vision, 2015, www.5g-ppp.eu
Understanding 5G: Perspectives on future technological advancements in mobile, 2014,
https://gsmaintelligence.com/research/?file=141208-5g.pdf&download
Kreutz D., Ramos F., Verissimo P., Rothenberg C.E, Azodolmolky S., Uhlig S. ‘Software-Defined
Networking: A Comprehensive Survey’. 2014, Available from http://arxiv.org/pdf/1406.0440.pdf
M.Mendonca, et. al., A Survey of Software-Defined Networking: Past, Present,and Future of
Programmable Networks, http://hal.inria.fr/hal-00825087/
ETSI GS NFV 003 V1.2.1 (2014-12), Network Functions Virtualization (NFV);Terminology for Main
Concepts in NFV,
http://www.etsi.org/deliver/etsi_gs/NFV/001_099/003/01.02.01_60/gs_NFV003v010201p.pdf
ETSI GS NFV 003 V1.2.1 (2014-12), Network Functions Virtualization (NFV);Terminology for Main
Concepts in NFV,
http://www.etsi.org/deliver/etsi_gs/NFV/001_099/003/01.02.01_60/gs_NFV003v010201p.pdf
ETSI GS NFV 002 v1.2.1 2014-12, NFV Architectural Framework
C.Kolias, Bundling NFV and SDN for Open Networking, NetSeminar @ Stanford, May 22, 2014,
http://netseminar.stanford.edu/seminars/05_22_14.pdf
C.Marcondes-NFV-WPEIF2014, https://www.google.ro/?gws_rd=ssl#q=ETSI+NFV+USe+cases++tutorial
J.Matias, J.Garay, N.Toledo, J.Unzilla, and E.Jacob,Toward an SDN-Enabled NFV Architecture, IEEE
Communications Magazine April 2015
ETSI -Network Functions Virtualization – Introductory White Paper,
https://portal.etsi.org/nfv/nfv_white_paper.pdf
Network Functions Virtualization – Update White Paper, https://portal.etsi.org/nfv/nfv_white_paper2.pdf
Network Functions Virtualization – White Paper #3,
https://portal.etsi.org/Portals/0/TBpages/NFV/Docs/NFV_White_Paper3.pdf
DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 100

References-2
17.
18.
19.
20.
21.
22.
23.

24.
25.
26.
27.
28.

M. Peng et al., “System Architecture and Key Technologies for 5G Heterogeneous Cloud Radio Access
Checko A., et al. 'Cloud RAN for Mobile Networks—A Technology Overview'. IEEE Communications
Surveys & Tutorials, VOL. 17, NO. 1, First Quarter 2015, 405
M. Peng, et al., “Heterogeneous cloud radio access networks: a new perspective for enhancing spectral
and energy efficiencies,” IEEE Wireless Commun., Dec. 2014, [online] available:
http://arxiv.org/abs/1410.3028.
Bradai A., Singh K., Ahmed T., and Rasheed T., ‘Cellular Software Defined Networking: A framework’.
IEEE Communications Magazine — Communications Standards Supplement, June 2015, pp. 36-43
Li L. E., Mao Z. M., and Rexford J.’Toward Software-Defined Cellular Networks’. European Workshop on
Software Defined Networking (EWSDN), 2012.
SK Telecom, ‘SK Telecom 5G White Paper: SK Telecom’s View on 5G Vision, Architecture, Technology,
and Spectrum’. Technical Report, 2014. Available from
http://www.sktelecom.com/img/pds/press/SKT_5G%20White%20Paper_V1.0_Eng.pdf
Granelli F., Gebremariam A.A., Usman M., Cugini F., Stamati V., Alitska M. and Chatzimisios P.,
‘Software Defined and Virtualized Wireless Access, in Future Wireless Networks: Scenarios and
Standards’. IEEE Communications Magazine —Communications Standards Supplement, June 2015,
pp.26 -33
China Mobile Research Institute, ‘CRAN White Paper: The Road Towards Green RAN’. October 2011,
Available from: http://labs.chinamobile.com/cran/wp-content/uploads/CRAN_white_paper_v2_5_EN.pdf
China Mobile Research Institute, ‘CRAN White Paper: The Road Towards Green RAN’. June 2014
Available from: http://labs.chinamobile.com/cran/wp-content/uploads/2014/06/20140613-CRAN-WP3.0.pdf
China Mobile Research Institute, ‘White Paper of Next Generation Fronthaul Interface’. Version 1.0’,
October 4, 2015, Available from http://labs.chinamobile.com/cran/2015/06/03/white-paper-of-ngfinextgeneration-fronthaul-interface-version-1-0/
Guerzoni R., Trivisonno R., Soldani D. ‘SDN-Based Architecture and Procedures for 5G Networks’. 5GU
2014, November 26-27, Levi, Finland, DOI 10.4108/icst.5gu.2014.258052
Julius M., et al. ‘Design and implementation of a Carrier Grade Software Defined Telecommunication
Switch and Controller’. Network Operations and Management Symposium (NOMS), IEEE, 2014

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 101

References-2
29.
30.
31.
32.
33.
34.
35.
36.
37.

Feng D., et.al. ‘Device-to-Device Communications in Cellular Networks’. IEEE Communications
Magazine, April 2014, pp.49-55
Akyildiz I.F., Wang P., Lin S.C. ‘SoftAir: A software defined networking architecture for 5G wireless
systems’, Computer Networks 85 (2015) pp.1–18.
Akyildiz I.F., Lin S., Wang P. ’Wireless software-defined networks(W-SDNs) and network function
virtualization (NFV) for 5G cellular systems: An overview and qualitative evaluation’. Computer Networks
93 (2015) 66–79, Available from http://dx.doi.org/10.1016/j.comnet.2015.10.013
Bernardos C.J., Oliva A., Serrano P., Contreras L.M., Banchs A., Jin H. ‘An architecture for software
defined wireless networking’, IEEE Wireless Communications Magazine,vol.21, no.3, June 2014, pp.5261
Guenter I. Klas, Fog Computing and Mobile Edge Cloud Gain Momentum Open Fog Consortium, ETSI
MEC and Cloudlets , Version 1.1 Nov 22, 2015
M.Chiang, "Fog Networking: An Overview on Research Opportunities“, December 2015,
https://arxiv.org/ftp/arxiv/papers/1601/1601.00835.pdf
Hung S.C, Hsu H. Lien S.Y., ChenK.C. ’Architecture Harmonization Between Cloud Radio Access
Networks and Fog Networks’. IEEE Access, December 2015. Available from
http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7359093
M.Patel et al., Mobile-Edge Computing – Introductory Technical White Paper,
https://portal.etsi.org/Portals/0/TBpages/MEC/Docs/Mobile-edge_Computing__Introductory_Technical_White_Paper_V1%2018-09-14.pdf
Yun Chao Hu et.al., "Mobile Edge Computing A key technology towards 5G" ETSI White Paper No. 11
September 2015, ISBN No. 979-10-92620-08-5

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 102

Backup slides

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 103

Network Function Virtualization
NFV Actors
ETSI NFV Group
Global (operators-initiated) Industry Specification Group (ISG) under the
auspices of ETSI
~200 members (2014)
‒28 Tier-1 carriers (and mobile operators) & service providers, cable industry
Open membership
ETSI members sign the “Member Agreement”
Non-ETSI members sign the “Participant Agreement”
Operates by consensus (formal voting only when required)
Deliverables: requirements specifications, architectural framework, PoCs, standards
liaisons
Face-to-face meetings quarterly.
Currently: four (4) WGs, two (2) expert groups (EGs), 4 root-level work items
(WIs)
WG1: Infrastructure Architecture
WG2: Management and Orchestration
WG3: Software Architecture
WG4: Reliability & Availability
EG1: Security
EG2: Performance &
Network Operators Council (NOC): technical advisory body

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 104

Network Function Virtualization
NFV Actors
Open Networking Foundation (ONF)
Active also in NFV area
E.g. of document: “OpenFlow-Enabled SDN and Network Functions
Virtualization,” 2014, see Refs.
Internet Research Task Force (IRTF)
RFC 7426, Jan 2015: “Software-Defined Networking (SDN): Layers and
Architecture Terminology” , see Refs.
proposes a common terminology for SDN layering and architecture based
on significant related work from the SDN research community

DataSys 2016 Conference May 22, 2016 Valencia, Spain
Slide 105