T u t o r ia l

SDN, NFV
a n d t h e i r r o l e i n 5G
Presenters
Katia Obraczka UC Santa Cruz, USA

Christian Rothenberg UNICAMP, Brazil

Ahmad Rostami Ericsson Research, Sweden

ACM SIGCOMM Tutorial | 2016-08-22 | Page 1
 A gen d a

Part I (9:00-10:30) Part II (11:00-12:30)

• 5G and the need for programmability • Research Efforts

• Software Defined Networking (SDN) • Use Cases
• Network Function Virtualization (NFV) • Q&A

Coffee Break at 10:30 for 30 Min.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 2

Outline
› Introduction to 5G: we will present a brief overview of 5G including the reference architecture,
opportunities, and challenges.
› SDN & NFV Concepts and their role in 5G: We will present SDN and NFV concepts and how they
enable the vision of future 5G networks. Concepts such as network programmability and abstractions
as well as network slicing will be introduced.
› SDN & NFV Enabling Technologies: SDN technologies for the control- and data planes will be
presented.. Notable examples of APIs for network abstraction and programmability, hardware
acceleration, as well as NFV-enabling features such as service function chaining and service
personalization will be covered.
› Related Projects and Use Cases: We will briefly overview SDN and NFV efforts that target future 5G
networks. We will also present use cases that can benefit from the synergy between SDN and NFV
including dynamic service chaining, such as virtualized telecommunication services (e.g., vCPE, vIMS,
vEPC, MEC) and NFV in radio access networks.
› Research Challenges and Future Directions: In addition to the challenges in realizing 5G networks,
we will further discuss research challenges posed by SDN and NFV including performance monitoring,
scalability, management and orchestration, security, portability and heterogeneous network support.
ACM SIGCOMM Tutorial | 2016-08-22 | Page 3
R e f e r e n c e s / L i t e r a t u r e (1/2)
› Nunes, B. A., Mendonca, M., Nguyen, X. N., Obraczka, K., & Turletti, T. (2014). A survey of software-defined
networking: Past, present, and future of programmable networks. IEEE Communications Surveys & Tutorials.
› Kreutz, D., Ramos, F. M., Esteves Verissimo, P., Esteve Rothenberg, C., Azodolmolky, S., & Uhlig, S. (2015).
Software-defined networking: A comprehensive survey. Proceedings of the IEEE.
› Rostami, A., "The Evolution of Programmable Networks: from Active Networks to Software Defined Networks
(SDN)," Tutorial presented at 26th International Teletraffic Congress (ITC), 2014.
› Mijumbi, R., Serrat, J., Gorricho, J. L., Latré, S., Charalambides, M., & Lopez, D. (2016). Management and
orchestration challenges in network functions virtualization. Communications Magazine, IEEE.
› S. Abdelwahab, B. Hamdaoui, M. Guizani and T. Znati, "Network function virtualization in 5G," in IEEE
Communications Magazine, vol. 54, no. 4, pp. 84-91, April 2016.
› Ericsson White paper, "5G system – enabling industry and society transformation," 2015.
https://www.ericsson.com/res/docs/whitepapers/what-is-a-5g-system.pdf
› Ericsson Mobility Report 2016. https://
www.ericsson.com/res/docs/2016/ericsson-mobility-report-2016.pdf
› Ericsson White Paper, “Cellular networks for massive IoT,”
https://www.ericsson.com/res/docs/whitepapers/wp_iot.pdf
› Ericsson Technology Review, “A vision of the 5G core:
flexibility for new business opportunities.”
›ACMEricsson White Paper, “Cloud RAN – the benefits of virtualization, centralization and coordination,”
SIGCOMM Tutorial | 2016-08-22 | Page 4
https://www.ericsson.com/res/docs/whitepapers/wp-cloud-ran.pdf
R e f e r e n c e s / L i t e r a t u r e (2/2)
› I. T. Haque; N. Abu-Ghazaleh. 2016. Wireless Software Defined Networking: a Survey and Taxonomy," in IEEE
Communications Surveys & Tutorials.
› Akram Hakiri and Pascal Berthou (2015) . Leveraging SDN for The 5G Networks: Trends, Prospects and
Challenges. http://arxiv.org/abs/1506.02876
› Nachikethas A. Jagadeesan and Bhaskar Krishnamachari. 2014. Software-Defined Networking Paradigms in
Wireless Networks: A Survey. ACM Comput. Surv.
› 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.
› 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
› 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
› Mao Yang, Yong Li, Depeng Jin, Lieguang Zeng, Xin Wu, and Athanasios V. Vasilakos. 2015. Software-Defined
and Virtualized Future Mobile and Wireless Networks: A Survey. Mob. Netw. Appl. 20, 1
› Nguyen, V., Do, T. & Kim, Y. SDN and Virtualization-Based LTE Mobile Network Architectures: A Comprehensive
Survey. Wireless Pers Commun (2016) 86: 1401. doi:10.1007/s11277-015-2997-7

ACM SIGCOMM Tutorial | 2016-08-22 | Page 5

5g a n d the need f o r
programmability
 Driving f o r c e s behind 5g

Networked Society Technical Drivers

• Improved User Experience • Network and Service
• Massive Traffic Volumes automation
• Massive No. of Connected • Resource & Energy
Devices Efficiency
• Massive No. of Services • Virtualization &
(e.g., IoT) Clouds
• Transformed Industries • New HW and SW
Technologies (SDN
& NFV)

ACM SIGCOMM Tutorial | 2016-08-22 | Page 7

 architecture

Devices/IoT Industry
Clouds
Management & Control

Access/Mobility Network Applications

Cloud/DC Infrastructure

Transport

ACM SIGCOMM Tutorial | 2016-08-22 | Page 8

source: Ericsson
 Global mobile t r a f f i c
(m onthly ExaBytes)

~45% CAGR

Growth in
12X
smartphone traffic

Around 90% of mobile traffic

will be from smartphones
by the end of 2021

ACM SIGCOMM Tutorial | 2016-08-22 | Page 9

source: Ericsson Mobility Report
IOT TO s u r p a s s m o b i l e p h o n e s i n 2 0 1 8

ACM SIGCOMM Tutorial | 2016-08-22 | Page 10

source: Ericsson Mobility Report
 E v o l u t i o n T o w a r d s 5G
1000x
Mobile 10x-100x
Data Connected
Volume Devices
5x
Lower
s
Laten
cy 10x-100x
End-user Data
Rates
10x
Battery Life
2G
~199
3G 4G 5G for
0 ~2000 ~2010 Low Power
ACM SIGCOMM Tutorial | 2016-08-22 | Page 11
~2020 Devices
Sour
ce:
MET
IS
 BROADBAND EXPERIENCE
EVERYWHERE, ANYTIME

5g
SMART VEHICLES,
TRANSPORT & INFRASTRUCTURE

MEDIA
EVERYWHERE

USE CRITICAL CONTROL

OF REMOTE DEVICES

CASES INTERACTIO
N HUMAN-
IOT

ACM SIGCOMM Tutorial | 2016-08-22 | Page 12 source: Ericsson 5G use

cases
 D i v e r s e R e q u i r e m e n t s o f IOT

ACM SIGCOMM Tutorial | 2016-08-22 | Page 13 source: Ericsson W.P. on cellular networks for massive IoT
One n e t w o r k – m u l t i p l e i n d u s t r i e s

Several Network Slices

ACM SIGCOMM Tutorial | 2016-08-22 | Page 14 source: Ericsson W.P. on 5G Systems
 Network a s a service
Network Service Catalog Service n

Mobile Wirelin Healt Network Service

Broadband e h
Orchestration
Health
Access Care
Nomadic Enterpris … … Robotic communication
Broadband e
… …
Premium Communication
Comm.
Industry Massive Media
Automatio Sensors/Act Mobile
n .
Broadband Basic
Service
Composition Network Slices

Resources Physical Resources (Access, Connectivity,

Connectivity NF
(Physical & Virtual) Computing, Storage, …)
Radio Cloud NF

ACM SIGCOMM Tutorial | 2016-08-22 | Page 15 source: Ericsson W.P. on 5G Systems

 Network Architecture
BBU Pool Cloud
EPC
BBU

Access Aggregation Core

Radio Access Transport Core Transport

Radio Packet Cloud Packet

uW Optical
Optical

ACM SIGCOMM Tutorial | 2016-08-22 | Page 16

 P r o g r a m m a b i l i t y i n 5G N e t w o r k s
High level of flexibility and programmability in individual domains (mobile core, radio access
network and transport network).
Cross-domain programmability and orchestration.

Service Agility Service Diversity Resource Efficiency

Shorten the time for Share a single Dynamically
service creation infrastructure among allocate the right
and service multiple services amount of
adaptation (e.g., with wide range of resources when
scaling). requirements. and where needed.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 17

 P r o g r a m m a b il it
y in Mobile
core
ACM SIGCOMM Tutorial | 2016-08-22 | Page 18
 C u r r e n t Mobile Core A r c h i t e c t u r e

• A single network architecture

for multiple services
MME HSS • Mix of control and user
plane functions
PCRF
PDN • Appliance-based realization
S-GW
GW

Evolved Packet Core (EPC)

• Difficult to customize
MME: Mobility Management Entity
S-GW: Serving Gateway
PDN: Packet Data Network
Internet • Scalability
PCRF: Policy & Charging Rules Function
HSS: Home Subscriber Server
ACM SIGCOMM Tutorial | 2016-08-22 | Page 19
 Flexible core architecture
• Separation of control and user-
plane functions
• Decompose core functionality into
granular functions
• Virtualize functions

• Customize realization per • Selective scaling

service/slice • Utilize Cloud Environment
• Centralized control • Flexible placement of functions
functions
Network Function Virtualization (NFV) is an enabler for programmability in mobile core.
ACM SIGCOMM Tutorial | 2016-08-22 | Page 20
Source of fig.: Ericsson Review on 5G Core
Flexibility
 P r o g r a m m a b il it
y i n RAN

ACM SIGCOMM Tutorial | 2016-08-22 | Page 21

 RAN D e p l o y m e n t s –
I
Distributed Baseband
Core
Networ
k • Flat Architecture
• Scaling
• IP connectivity between RAN
Backhaul
and Core, and among sites

L3

L2
ACM SIGCOMM Tutorial | 2016-08-22 | Page 22

L1
 RAN D e p l o y m e n t s –
II
Centralized Baseband (C-RAN)
Core
• Pooling gains
Networ
k
• Efficient network management
• Efficient coordination &
Backhaul
interference management
BBU • Less network signaling
Pool
Fronthaul • Stringent performance requirement on
CPRI
fronthaul (BW, delay and jitter)
• might not be scalable in all 5G
scenarios
Need for more flexible split of
RAN
ACM SIGCOMM Tutorial | 2016-08-22 | Page 23 CPRI: Common Public Radio BBU: Baseband processing Unit
Interface
 CLOUD RAN

Centralized Split 1 Split 2 Distributed

L3 L3 L3
L2 L2
L3
L1
L2 L2
L1
L1 L1

ACM SIGCOMM Tutorial | 2016-08-22 | Page 24 Source of fig.: Ericsson W.P. on Cloud RAN
 Flexibility with c l o u d r a n

Centralization gains as in
L3 C-RAN
L2 high • Pooling
• Network Management
L2 low
• Coordination
L1
Less Transport Requirements
L3
L2
L1 Virtualization gains
• Selective scaling (E.g.
User plane vs Control
Plane)
• Cloud-based
Realization
L3 Collocation of RAN & Core
L2
ACM SIGCOMM Tutorial | 2016-08-22 | Page 25
L1 Source of fig.: Ericsson W.P. on Cloud RAN
 P r o g r a m m a b il it
y in Transport

ACM SIGCOMM Tutorial | 2016-08-22 | Page 26

 Current transport networks

Access Aggregation Core

Packet Packet Packet

uW uW
Optical Optical Optical

ACM SIGCOMM Tutorial | 2016-08-22 | Page 27

Current t r a n s p o r t n e t w o r k ISsues

• Monolithic realization of
control and forwarding • Lengthy and manual
functions service creation/scaling
• Proprietary Management • Inefficient Resource
Interfaces Utilization
• Complex control • Inefficient static
and management sharing
• Several technology • Difficult cross-layer
domains with optimization
independent control • Application
Unaware
ACM SIGCOMM Tutorial | 2016-08-22 | Page 28
 Progr am m abl e Transport
• Separation of control • Automation of network and
and forwarding functions services
• Define interfaces • Dynamic creation/update of
between control and (virtual) connections/tunnels
• Resource-optimized
forwarding
operation
• Open up the control • Cross-layer optimization
plane for programming (e.g. packet-optical
• Develop Efficient sharing convergence)
mechanisms • Radio-aware
adaptations

Software-Defined Networking (SDN) is an enabler for programmability in transport networks.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 29
S o f t w a r e - D ef in e d
Networking

Towards Network Programmability

Outline

› What’s Software-Defined Networking?

› Why Software-Defined Networking?
› Some History
› SDN Architectures
› Definitions, Terminology, Concepts
› Example Applications

ACM SIGCOMM Tutorial | 2016-08-22 | Page 31

 W h a t ’s S o f t w a r e - D e f i n e d
Networking?

ACM SIGCOMM Tutorial | 2016-08-22 | Page 32

 N e t w o r k ’s F u n c t i o n a l P l a n e s

Source: “Software-Defined Networking: A Comprehensive Survey”, Kreutz et al., https://arxiv.org/pdf/1406.0440.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 33
 W h a t ’s S o f t w a r e - D e f i n e d
Networking?
› Main principle: data plane decoupled from control plane.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 34

 Why S o f t w a r e - D e f i n e d
Networking?
› The Internet has been
the victim of its own
success!
› Extremely hard to
configure, manage, and
evolve.
› “Vertically integrated”: tight
coupling of control- and data
planes
embedded/distributed in
network devices.
› Proliferation of
“middleboxes”.
ACM SIGCOMM Tutorial | 2016-08-22 | Page 35
Source: N. McKeown, Stanford, ONF
Why S o f t w a r e - D e f i n e d
Networking?
• The Internet has been the Software-Defined Networking to the
victim of its own rescue!
success! › Separation of control plane from data
• Extremely hard to configure, plane.
manage, and evolve. – Network control “logically centralized” in
• “Vertically integrated”: tight the controller.
– Forwarding hardware simplified.
coupling of control- and
data planes › Programmable networks to facilitate
embedded/distributed in management and control and combat
network devices. “network ossification”.
• Proliferation of › Data plane “commoditization”.

“middleboxes”
.
ACM SIGCOMM Tutorial | 2016-08-22 | Page 36
Software-Defined Networking:
Some History

ACM SIGCOMM Tutorial | 2016-08-22 | Page 37

 P r o g r a m m a b l e Networks: Some
History

1990’s

Open Devolved
Active
Signaling Control of
Networking
(OPENSIG) ATM (DCAN)

General
Switch
Management Capsules
Protocol User-
Program
(GSMPv3, RFC 3292, 2002) mable
Switches

ACM SIGCOMM Tutorial | 2016-08-22 | Page 38

 P r o g r a m m a b l e Networks: Some
History

2000’s

IETF
4D Project ETHANE
NETCONF

Separation SDN
between SNMP predecessor
routing and successor
data plane

ACM SIGCOMM Tutorial | 2016-08-22 | Page 39

 Software-Defined Networking
Architectures
IETF ForCES ONF OpenFlow
› Forwarding Element (FE) and Control › Decoupling between control- and
Element (CE) data planes
› Both reside in the network device › Controller and switch communicate
› FE and CE communicate using the using the OpenFlow protocol
ForCES protocol

ACM SIGCOMM Tutorial | 2016-08-22 | Page 40

 ONF O p e n F l o w A r c h i t e c t u r e

ACM SIGCOMM Tutorial | 2016-08-22 | Page 41

 ONF O p e n F l o w A r c h i t e c t u r e
Controller
OpenFlow Switch specification

OpenFlow Switch PC

Secure
sw
Channe
l

hw Flow IANA port for OpenFlow Switch-

Controller connection: 6653
Tabl
e

Source: The Stanford Clean Slate Program,

ACM SIGCOMM Tutorial | 2016-08-22 | Page 42 http://cleanslate.stanford.edu
O p e n F l o w 1.0 F l o w T a b l e & F i e l d s
Ingres Ethernet VLAN IP TCP/UDP
Header Fields s SA DA Type ID Priority SA DA Proto TOS Src Dst
Port

Classifier Action Statistics

Classifier Action Statistics
Flow Table Classifier Action Statistics
OF1.0 style

…
Classifier Action Statistics

Physical Port
ALL
CONTROLLER
Forward Virtual
Actions Port LOCAL
TABLE
IN_PORT
Mandatory Action

Optional Action

Drop
Virtual NORMAL
Forward
Port FLOOD
ACM SIGCOMM Tutorial | 2016-08-22 | Page 43
Enqueue
Modify Field
 O p e n F l o w Ta b l e E n t r i e s

Each Flow Table entry has two timers:

hard_timeout idle_timeout
seconds after which the flow is seconds of no matching packets
removed after which the flow is removed
zero mean never times-out zero means never times-out

If both idle_timeout and hard_timeout are set, then the flow is removed when the first of the two
expires.
ACM SIGCOMM Tutorial | 2016-08-22 | Page 44
 OpenFlow Standards

Evolution path:
• OF 1.0 (03/2010): Most widely used version, MAC, IPv4, single table (from Stanford)
• OF 1.1 (02/2011): MPLS tags/tunnels, multiple tables, counters (from Stanford)
• OF 1.2 (12/2011): IPv6, extensible expression
• OF-Config 1.0 (01/2012): Basic configuration: queues, ports, controller assign
• OF 1.3.0 (04/2012): Tunnels, meters, PBB support, more IPv6
• OF-Config 1.1 (04/2012): Topology discovery, error handling
• OF-Test 1.0 (2H2012): Interoperability conformance test processes, suites, labs
• OF 1.3.2 (May 2013), 19 errata, final review
• OF 1.4 (Aug. 2013), 9 changes + 13 extensions, More extensible wire protocol,
Flow monitoring, Eviction, Vacancy events, Bundles
• OF 1.5.1 (Dec. 2014), Egress Tables, Packet type aware pipeline, Extensible flow entry
statistics
ACM SIGCOMM Tutorial | 2016-08-22 | Page 45 Source: ONF
 SDN: S o m e D e f i n i t i o n s

ACM SIGCOMM Tutorial | 2016-08-22 | Page 46

 SDN: S o m e D e f i n i t i o n s

› “The SDN architecture decouples the network control and forwarding functions
enabling the network control to become directly programmable and the
underlying infrastructure to be abstracted for applications and network services.”
Open Networking Foundation (opennetworking.org)

› “Software Defined Networking (SDN) refactors the relationship between network

devices and the software that controls them. Opening up the interfaces to
programming the network enables more flexible and predictable network
control, and makes it easier to extend the network with new functionality.” ACM
Sigcomm Simposium on Software-DefinedNetworking Research 2016

ACM SIGCOMM Tutorial | 2016-08-22 | Page 47

SDN: D e f i n i t i o n s , C o n c e p t s , a n d
Te r m i n o l o g y
SDN refers to software-defined networking SDN architecture
architectures where:
• Data- and control planes decoupled from one
another.
• Data plane at forwarding devices managed and
controlled remotely by a “controller”.
• Well-defined programming interface between
control- and data planes.
• Applications running on controller manage and
control underlying data plane

Source:
“Software-Defined Networking: A Comprehensive Survey”,
Kreutz et al., https://arxiv.org/pdf/1406.0440.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 48

SDN: D e f i n i t i o n s , C o n c e p t s , a n d
Te r m i n o l o g y
› Data plane: network infrastructure consisting of
interconnected forwarding devices (a.k.a., forwarding
plane).
› Forwarding devices: data plane hardware- or software
devices responsible for data forwarding.
› Flow: sequence of packets between source-destination
pair; flow packets receive identical service at forwarding
devices.
› Flow rules: instruction set that act on incoming packets
(e.g., drop, forward to controller, etc)
› Flow table: resides on switches and contains rules to
handle flow packets.

Source:
“Software-Defined Networking: A Comprehensive Survey”,
Kreutz et al., https://arxiv.org/pdf/1406.0440.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 49

SDN: D e f i n i t i o n s , C o n c e p t s ,
a n d Te r m i n o l o g y
› Control plane: controls the data
plane; logically centralized in
the “controller” (a.k.a., network
operating system).

› Southbound interface:
(instruction set to program the
data plane) + (protocol between
control- and data planes).
Source:
“Software-Defined Networking: A Comprehensive Survey”,
Kreutz et al., https://arxiv.org/pdf/1406.0440.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 50

SDN: D e f i n i t i o n s , C o n c e p t s , a n d
Te r m i n o l o g y
› Northbound interface: API
offered by control plane to
develop network control- and
management applications.
› Management plane: functions,
e.g., routing, traffic
engineering, that use control
plane functions and API to
manage and control
network infrastructure. Source:
“Software-Defined Networking: A Comprehensive Survey”,
Kreutz et al., https://arxiv.org/pdf/1406.0440.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 51

 SDN A p p l i c a t i o n s
› Traffic engineering: › Measurement and monitoring:
– Provide adequate QoS, improve – Packet sampling, traffic matrix
network utilization, reduce power estimation
consumption, balance load. › Security:
› Wireless network – Firewalling, access control, DoS
management/control and mobility attack detection/ mitigation, traffic
support: anomaly detection.
– Seamless handover, load balancing, › Data-center networking:
interoperability between – Data center QOS and traffic
heterogeneous networks, dynamic engineering, fault detection and
spectrum usage. resilence, dynamic provisioning,
security.

Source: “Software-Defined Networking: A Comprehensive Survey”, Kreutz et al., https://arxiv.org/pdf/1406.0440.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 52
ONF r e c u r s i v e SCDusNtocmonetrroRllearppRlicatio

SDN a r c h i t e c t u r e n
Controller plane
Manager
R
(
P
h
y
s
Controller plane (Virtual) data plane
Manager i
G c
(Physical) data plane a
l
)
SDN controller B
Controller plane (Virtual) datad plane (Virtual) data plane
Manager a
B t
(Physical) data plane Source: ONF a TR-504 : SDN Architecture Overview Version 1.1, https://
www.opennetworking.org/images/stories/downloads/sdn-
p
resources/technical-reports/TR_SDN-ARCH-Overview-1.1-
ACM SIGCOMM Tutorial | 2016-08-22 | Page 53 11112014.02.pdf
l
a
n
e
 Network Function Virtualization
(NFV)

ACM SIGCOMM Tutorial | 2016-08-22 | Page 54

 Motivation

Source: Why Virtualization is Essential for 5G – Francis Chow (5G Summit 2015)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 55
 Problem Statement

• Complex carrier networks

– with a large variety of proprietary nodes and hardware appliances
• Launching new services is difficult and takes too long
• Space and power to accommodate
– requires just another variety of box, which needs to be integrated
• Operation is expensive
– Rapidly reach end of life due to
• existing procure-design;
• integrate-deploy cycle

ACM SIGCOMM Tutorial | 2016-08-22 | Page 56

 Some Changes

Source: ETSI NFV ISG – DIRECTION & PRIORITIES – Steven Wright (NFV World Congress 2015)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 57
 Transformation

ACM SIGCOMM Tutorial | 2016-08-22 | Page 58

Source: Adapted from D. Lopez Telefonica I+D, NFV
 Why NFV?

1.Virtualization: Use network resource without worrying about where it is physically

located, how much it is, how it is organized, etc.
2. Orchestration: Manage thousands of devices
3. Programmable: Should be able to change behavior on the fly.
4. Dynamic Scaling: Should be able to change size, quantity, as a F(load)
5. Automation: Let machines / software do humans’ work
6. Visibility: Monitor resources, connectivity
7. Performance: Optimize network device utilization
8. Multi-tenancy: Slice the network for different customers (as-a-Service)
9. Service Integration: Let network management play nice with OSS/BSS
10. Openness: Full choice of modular plug-ins
Note: These are exactly the same reasons why we need/want SDN.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 59

 The NFV Concept
A means to make the network more flexible and simple by
minimising dependence on HW constraints

ACM SIGCOMM Tutorial | 2016-08-22 | Page 60

Source: Adapted from D. Lopez Telefonica I+D, NFV
 The Making of NFV

Souce: NFV Orchestration | Fueling innovation in operator networks - Federico

D2 e| Psacgea6l1zo (TM FORUM 2016)
ACM SIGCOMM Tutorial | 2016-08-2
 Benefits & Promises of NFV

• Reduced equipment costs (CAPEX)

– through consolidating equipment and economies of scale of IT industry.
• Increased speed of time to market
– by minimising the typical network operator cycle of innovation.
• Availability of network appliance multi-version and multi-tenancy,
– allows a single platform for different applications, users and tenants.
• Enables a variety of eco-systems and encourages openness.
• Encouraging innovation to bring new services and generate new
revenue streams.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 62 Source: NFV

 Benefits & Promises of NFV

• Flexibility to easily, rapidly, dynamically provision and

instantiate new services in various locations
• Improved operational efficiency
• by taking advantage of the higher uniformity of the physical network
platform and its homogeneity to other support platforms.
• Software-oriented innovation to rapidly prototype and test
new services and generate new revenue streams
• More service differentiation & customization
• Reduced (OPEX) operational costs: reduced power, reduced
space, improved network monitoring
• IT-oriented skillset and talent

ACM SIGCOMM Tutorial | 2016-08-22 | Page 63

Source: Adapted from D. Lopez Telefonica I+D, NFV
 ETSI NFV Architectural Framework

S o uPacgee6:4 ETSI NFV White

ACM SIGCOMM Tutorial | 2016-08-2 2 |

Paper 2
 NFV

Source: View on 5G Architecture - 5G PPP Architecture Working Group (2016)

ACM SIGCOMM Tutorial | 2016-08-22 | Page 65
 NFV Layers
E2E Network Service
End End
Poin Network Service Poin
t t
Logical
Abstractions VNF

Logical Links
VNF
VNF VNF VNF
VNF Instances
SW Instances
VNF VNF VNF VNF
VNF : Virtualized Network Function

NFV Infrastructure

Virtual Resources Virtual Virtual Virtual

ComputeStorage
Network
Virtualization SW
Virtualization Layer
HW Resources Compute Storage Network

ACM SIGCOMM Tutorial | 2016-08-22 | Page 66

Source: Adapted from D. Lopez Telefonica I+D, NFV
 Rethinking relayering

ACM SIGCOMM Tutorial | 2016-08-22 | Page 67

 NFV Concepts
• Network Function (NF): Functional building block with a well defined interfaces and well defined
functional behavior
• Virtualized Network Function (VNF): Software implementation of NF that can be deployed in a
virtualized infrastructure
• VNF Set: Connectivity between VNFs is not specified,
e.g., residential gateways
• VNF Forwarding Graph: Service chain when network connectivity order is important, e.g., firewall,
NAT, load balancer
• NFV Infrastructure (NFVI): Hardware and software required to deploy, mange and execute VNFs
including computation, networking, and storage.
• NFV Orchestrator: Automates the deployment, operation, management, coordination of VNFs and
NFVI.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 68

Source: Adapted from Raj Jain
 NFV Concepts
• NFVI Point of Presence (PoP): Location of NFVI
• NFVI-PoP Network: Internal network
• Transport Network: Network connecting a PoP to other PoPs or external networks
• VNF Manager: VNF lifecycle management e.g., instantiation, update, scaling, query, monitoring, fault
diagnosis, healing, termination
• Virtualized Infrastructure Manager: Management of computing, storage, network, software
resources
• Network Service: A composition of network functions and defined by its functional and behavioral
specification
• NFV Service: A network services using NFs with at least one VNF.

ACM SIGCOMM Tutorial | 2016-08-22 | Page 69

Source: Adapted from Raj Jain
 NFV Concepts
• User Service: Services offered to end users/customers/subscribers.
• Deployment Behavior: NFVI resources that a VNF requires, e.g., Number of VMs, memory, disk,
images, bandwidth, latency
• Operational Behavior: VNF instance topology and lifecycle operations, e.g., start, stop, pause,
migration, …
• VNF Descriptor: Deployment behavior + Operational behavior

ACM SIGCOMM Tutorial | 2016-08-22 | Page 70

Source: Adapted from Raj Jain
 Overview of ETSI NFV Use Cases

Souce: ETSI NFV White Paper 2

ACM SIGCOMM Tutorial | 2016-08-22 | Page 71
 Architectural Use Cases
• Network Functions &Virtualisation Infrastructure as a Service
– Network functions go cloudlike
• Virtual Network Function as a Service
– Ubiquitous, delocalized network functions
• Virtual Network Platform as a Service
– Applying multi-tenancy at the VNF level
• VNF Forwarding Graphs
– Building E2E services by composition

ACM SIGCOMM Tutorial | 2016-08-22 | Page 72

 Service-Oriented Use Cases
• Mobile core network and IMS
– Elastic, scalable, more resilient EPC
– Specially suitable for a phased approach
• Mobile base stations
– Evolved Cloud-RAN
– Enabler for SON
• Home environment
– L2 visibility to the home network
– Smooth introduction of residential services
• CDNs
– Better adaptability to traffic surges
– New collaborative service models
• Fixed access network
– Offload computational intensive optimization
– Enable on-demand access services
ACM SIGCOMM Tutorial | 2016-08-22 | Page 73
 NFV Framework Requirements
1. General: Partial or full Virtualization, Predictable performance
2. Portability: Decoupled from underlying infrastructure
3.Performance: Conforming and proportional to NFs specifications and facilities to
monitor
4. Elasticity: Scalable to meet SLAs. Movable to other servers.
5. Resiliency: Be able to recreate after failure.
Specified packet loss rate, calls drops, time to recover, etc.
6. Security: Role-based authorization, authentication
7. Service Continuity: Seamless or non-seamless continuity after failures or
migration

ACM SIGCOMM Tutorial | 2016-08-22 | Page 74

 NFV Framework Requirements
8.Service Assurance: Time stamp and forward copies of
packets for Fault detection
9.Energy Efficiency Requirements: Should be possible to put a subset of VNF in a power
conserving sleep state
10.Operational and Management Requirements: Incorporate mechanisms for automation
of operational and management functions
11.Transition: Coexistence with Legacy and Interoperability among multi-vendor
implementations
12.Service Models: Operators may use NFV infrastructure
operated by other operators

...

ACM SIGCOMM Tutorial | 2016-08-22 | Page 75

 Challenging Path upfront:
Not as simple as cloud applied to telco
The network differs from the computing environment in
2 key factors…
1
Data plane workloads HIGH PRESSURE ON
(which are huge!) PERFORMANCE

2
Network requires shape GLOBAL NETWORK VIEW IS
(+ E2E interconnection) REQUIRED FOR
MANAGEMENT

…which are big challenges for vanilla cloud computing.

AN ADAPTED VIRTUALISATION ENVIRONMENT IS NEEDED
TO OBTAIN CARRIER-CLASS BEHAVIOUR

ACM SIGCOMM Tutorial | 2016-08-22 | Page 76

Source: Adapted from D. Lopez Telefonica I+D, NFV
 The Road to NFV

Source: Gabriel Brown, Heavy Reading

ACM SIGCOMM Tutorial | 2016-08-22 | Page 77
 Fat vs. fit VNFs

Source: SDN and NFV Stepping Stones to the Telco Cloud – Prodip Sen (ONS 2016)

ACM SIGCOMM Tutorial | 2016-08-22 | Page 78

 Alternative options to virtualize NFV apps

ACM SIGCOMM Tutorial | 2016-08-22 | Page 79

 Performance Challenges

Souce: ETSI NFV White Paper 2

ACM SIGCOMM Tutorial | 2016-08-22 | Page 80
 Portability Challenges

Source: Adapted from D. Lopez Telefonica I+D, NFV

ACM SIGCOMM Tutorial | 2016-08-22 | Page 81
 Integration Challenges

Source: Adapted from D. Lopez Telefonica I+D, NFV

ACM SIGCOMM Tutorial | 2016-08-22 | Page 82
 Elasticity Challenges

ACM SIGCOMM Tutorial | 2016-08-22 | Page 83 Source: UNIFY Project Results

 Management & Orchestration Challenges

ACM SIGCOMM Tutorial | 2016-08-22 | Page 84 Source: IEEE PIMRC, 2013

 NFV & SDN

Source: SDN and NFV Stepping Stones to the Telco Cloud – Prodip Sen (ONS 2016)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 85
 SDN & NFV
• SDN and NFV do NOT depend on each other

ACM SIGCOMM Tutorial | 2016-08-22 | Page 86 Source: Uwe Michel, T-Systems

 NFV vs SDN
SDN ››› flexible forwarding &
steering of traffic in a physical or
virtual network environment
[Network Re-Architecture]

NFV ››› flexible placement of

virtualized network functions across
the network & cloud
[Appliance Re-Architecture]
(initially)

››› SDN & NFV are complementary

tools for achieving full network
programmability

ACM SIGCOMM Tutorial | 2016-08-22 | Page 87

 Flexibility with SDN & NFV

ACM SIGCOMM Tutorial | 2016-08-22 | Page 88 Source: Ahmad Rostami, Ericsson Research (Kista): http://www.itc26.org/fileadmin/ITC26_files/ITC26-Tutorial-Rostami.pdf
 SDN & NFV Convergence

ACM SIGCOMM Tutorial | 2016-08-22 | Page 89

Source: SDN and NFV Stepping Stones to the Telco Cloud – Prodip Sen (ONS 2016)
 5G + (NFV & SDN)

ACM SIGCOMM Tutorial | 2016-08-22 | Page 90

Source: View on 5G Architecture - 5G PPP Architecture Working Group (2016)
 5G Softwarization and Programmability Framework

ACM SIGCOMM Tutorial | 2016-08-22 | Page 91 Source: View on 5G Architecture - 5G PPP Architecture Working Group (2016)
 Multi Administrative Domains

ACM SIGCOMM Tutorial | 2016-08-22 | Page 92 Source: View on 5G Architecture - 5G PPP Architecture Working Group (2016)
 The Big Picture

ACM SIGCOMM Tutorial | 2016-08-22 | Page 93 Source: View on 5G Architecture - 5G PPP Architecture Working Group (2016)
 SDN/NFV Open Innovation

ACM SIGCOMM Tutorial | 2016-08-22 | Page 94 Source: SDN and NFV Stepping Stones to the Telco Cloud – Prodip Sen (ONS 2016)
 NFV enables MEC: Mobile Edge Computing
• MEC provides IT and cloud-computing capabilities within the RAN in close
proximity to mobile subscribers to accelerate content, services and applications
so increasing responsiveness from the edge .
• Standardization bodies: ETSI, 3GPP, ITU-T
• RAN edge offers a service environment with ultra-low latency and high
bandwidth 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)
• Source: https://portal.etsi.org/Portals/0/TBpages/MEC/Docs/Mobile-
edge_Computing_-_Introductory_Technical_White_Paper_V1%2018-09-14.pdf
ACM SIGCOMM Tutorial | 2016-08-22 | Page 95
 MEC:
Mobile
Edge
Computing

Deployment scenarios of the

Mobile-edge Computing server

ACM SIGCOMM Tutorial | 2016-08-22 | Page 96

• Source: https://portal.etsi.org/Portals/0/TBpages/MEC/Docs/Mobile-edge_Computing_-_Introductory_Technical_White_Paper_V1%2018-09-14.pdf
 MEC:
Mobile
Edge
Computing

MEC server platform overvie w

ACM SIGCOMM Tutorial | 2016-08-22 | Page 97

• Source: https://portal.etsi.org/Portals/0/TBpages/MEC/Docs/Mobile-edge_Computing_-_Introductory_Technical_White_Paper_V1%2018-09-14.pdf
 Summing Up

Source: Software-defined networking (SDN): a Dell point of view - A Dell White Paper (2015)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 98
 Enabling Technologies &
Open Source Efforts

ACM SIGCOMM Tutorial | 2016-08-22 | Page 99

 Enabling Technologies
• Virtualization & Minimalistic OS
– Docker, ClickOS, Unikernel
• Improving Linux I/O &
x86 for packet processing
– DPDK, Netmap, VALE,
Linux NAPI
• Programmable virtual
– Open vSwitch, P4, Open Networking Linux Image source: ClickOS
• switches / bridges
Example start-ups
– LineRate Systems, 6WIND, Midonet, Vyatta (bought by BCD)

Image source: NEC

ACM SIGCOMM Tutoria l | 2016-08-22 | Page 100

 Enabling Technologies: Open Source
Why Open Source in Networking?
• Higher reliability, more flexibility
• Faster, lower cost, and higher quality development
• Collaborative decisions about new features and roadmaps
• A common environment for users and app developers
• Ability for users to focus resources on differentiating development
• Opportunity to drive open standards

Bottom Line: The open source model significantly accelerates

consensus, delivering high performing, peer-reviewed code that
forms a basis for an ecosystem of solutions.
Source: Open Source in a Closed Network – Prodip Sen (OPNFV Summit 2015)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 101
 SDN/NFV & Open Source: Evolving and
accelerating the path of standardization

Further Reading:
• IETF Trends and Observations draft-arkko-ietf-trends-and-observations-00
• Source of table: "When Open Source Meets Network Control Planes." In IEEE Computer
(Special Issue on Software-Defined Networking), vol.47, no.11, pp.46,54, Nov. 2014.
• Source of figure: A. Manzalini et al., “Towards 5G Software-Defined Ecosystems”
ACM SIGCOMM Tutorial | 2016-08-22 | Page 102
 Standard / Open Source Organizations

Industry

Academia

Source: SDN IEEE Outreach, http://sdn.ieee.org/outreach

ACM SIGCOMM Tutorial | 2016-08-22 | Page 103
 Foundations
Target Collaboration
• Neutral and non-competing
• Legal framework for licensing, copyright, intellectual property management

"Companies feel they can collaborate on an open source project through an independent, not-for-profit entity that they trust
- this is incredibly important to them," --Allison Randa (Board President of Open Source Initiative)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 104
 Open Source Building Blocks
2015 – 2016: Several New Projects

ACM SIGCOMM Tutorial | 2016-08-22 | Page 105 Source: The Open Source NFV Eco-system and OPNFV’s Role Therein – Frank Brockners (OPNFV Summit 2016)
 Open Sourrce SDN Projects (2014 snapshot)

Source: "When Open Source Meets Network Control Planes." In IEEE Computer (Special Issue on Software-Defined Networking). 2014.
ACM SIGCOMM Tutorial | 2016-08-22 | Page 106
 NEW: P l e a s e s e e
and contribute to
› https://goo.gl/XCGDGS
› http://bit.do/oss-sdn-nfv

https://docs.google.com/spreadsheets/d/1NHI4MZZWVDpxF_Rs7OOSTUa_aHL2ACUVA_Ov-YQs1DA/edit#gid=0
ACM SIGCOMM Tutorial | 2016-08-22 | Page 107
 A growing ecosystem...

ACM SIGCOMM Tutorial | 2016-08-22 | Page 108

 5G Related Open Source

ACM SIGCOMM Tutorial | 2016-08-22 | Page 109

 Experimentation

ACM SIGCOMM Tutorial | 2016-08-22 | Page 110

 Testbed
s

ACM SIGCOMM Tutorial | 2016-08-22 | Page 111 Source: http://www.5g-berlin.org/ http://openfederatedtestbed.org/

 Continuous Integration Environment
• Continuous Integration tool chain speeds development and
facilitates collaboration
– Gerrit – code review tool
– GitHub - code repository
– Jenkins – automated build tool
– Maven – code build
– Ansible/Chef/Puppet – code deployment tool
– Docker/Vagrant – deployment to Containers/VMs

Source: Open Source Carrier Networking – Chris Donley (OPNFV Summit 2016)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 112
 Example: NTT (TM FORUM 2016)

Source: A Transformation From Legacy Operation to Agile Operation – Makoto Eguchi (TM
2F1
ACM SIGCOMM Tutorial | 0 6O80-R2-U| MgP
a2e 101316)
 The Frontier of Networking

Adapted from: Kyle Mestery, Next Generation Network Developer Skills

ACM SIGCOMM Tutorial | 2016-08-22 | Page 114
 Some Takeaways
• Open source speeds up development and promotes interoperability
• Rapid prototypes are a great way to show SDN/NFV value, and now 5G too!
• Developer ramp-up time is challenging,
– but once integrated, we can make rapid progress
• A common CI environment facilitates sharing between projects
• Cadence, short iteration cycles, fast feedback
• Some people are saying:
– “open source is the new standardization”
– “code is the coin of the realm”
• Stay tuned:
– http://www.sdn-os-toolkits.org/
– http://sdn.ieee.org/outreach

ACM SIGCOMM Tutorial | 2016-08-22 | Page 115

 Thank you!

Questions?

ACM SIGCOMM Tutorial | 2016-08-22 | Page 116

 BACKUP

ACM SIGCOMM Tutorial | 2016-08-22 | Page 117

 Perspective

Source: Survey Results: Bridging the Gap Between Open Standards and Open Source - Elizabeth Rose
(OPNFV Summit 2016)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 118
 Challenges: Closed vs. Open

Source: Open Source in a Closed Network – Prodip Sen (OPNFV Summit 2015)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 119
 Open Source Road

Source: The NFV Revolution Must Be Open – Dave Neary (OPNFV Summit 2016)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 120
 Approach for Contributions
• Open Source Projects need more • Evaluate how your project idea
than code! can fit with the existing
– Documentation project
– QA – Does it overlap?
– Infrastructure – Does it provide extra value?
– Blogs – Can something be abstracted?
– IRC – Is the community interested?
– Does it fit the communities
goals?
– Do you have a plan for testing,
documentation and support?
Source: Upstream Open Source Networking Development: The Good, The Bad, and the Ugly – Kyle Mestery,
J u s t i n P e t i t , Re 1u21ssell Bryant (ONS 2016)
ACM SIGCOMM Tutoria l | 2 0 1 6 -08 -2 2 | P a g
 Project Evolution: Examples

Source: Upstream Open Source Networking Development: The Good, The Bad, and the Ugly – Kyle Mestery,
J u20s1t6i-n08-P22et| tPiatg,eR12u2ssell Bryant (ONS 2016)
ACM SIGCOMM Tutorial |
 Building on Open Source Projects

“Do NOT fall into the trap of adding

more and more stuff to an out-of-
tree project. It just makes it harder
and harder to get it merged. There
are many examples of this.”
-Andrew Morton
Source: Swimming Upstream – Dave Neary (OPNFV Summit 2016)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 123
 Open Ended Questions
• What are the inhibitors to adoption of Open Source Implementations?
– Issues surrounding licensing (esp. GPL)
– Industry understanding of Open Source licensing
– Competitive issues and fragmentation
– Security
– Quality and robustness
– Maintenance and support

Source: Survey Results: Bridging the Gap Between Open Standards and Open Source - Elizabeth Rose
(OPNFV Summit 2016)
ACM SIGCOMM Tutorial | 2016-08-22 | Page 124