Market Outlook
EPC investment
wave #2 starts now
Operators have
Subscriber and session growth continues
Voice and video over LTE Operator
’s
Operators
have
started investing in
EPC capacity to
cope with signaling
and traffic
g
rowth in
continues Machine to Machine growth driving huge EPC capacity needs for o
p
erators in Asia
,
Operator s
investment on
experience mgmt,
voice capability and
video o
p
timization
g
4G network
p,
Europe and America
p
are essential
Disruption with cloud Disruption
Packet Core Portfolio
Radio Access
HSS
Evolved Packet Core
Gb
Anchor point for subscriber mobility
between Radio Access
Packet Core is the
anchor point for
subscriber mobility
Operator Voice and SMS
Services
PCRF
IMS/VoLTE
S
1
MME
S11
S6a
Gx
SGi
Rx
BTS
NodeB
RNC
BSC
Iu
SGSN/MME
(Flexi NS)
subscriber
mobility
between radio access
and services, and
control point for quality
of service and traffic
monetization
Mobile Broadband
eNodeB
Corporate Connectivity
Service
GGSN/S/P-GW
(Flexi NG)
S
1
-
MME
SGi
Control point for quality of service and
traffic monetization
monetization
Mobile
Functional Units MME
Architecture
MMDU: Mobility Management & Database Unit •
Subscriber database and subscriber state based mobility
management for the MME
CPPU: Control Plane Processing Unit
OMU
(2n)
CPPU
(n+)
CPPU:
Functional Units SGSN
Architecture
VMU: Virtualization Management Unit •
Hosts logical PAPU and SMMU. PAPU processes 2G and 3G control
planes. The SMMU handles mobility management operations
MMDU: Mobility Management & Database Unit
•
Subscriber database
OMU
(2n)
CPPU
(n+)
•
Subscriber
database
IPDU: IP Director Unit •
Load balancing, connectivity and IP address virtualization
IPPU: IP Packet Processing Unit •
3G user
p
lane throu
g
h
p
ut handlin
g
Ethernet Ethernet
MCHU
(2n)
VMU
(n+)
PAPU
(n+1)
SMMU
(n+1)
pgpg
GBU: Gb Interface Unit •
2G user plane throughput handling
CABU: Carrier Blade Unit •
Houses Frame Relay Interface Units (FRIU). External STM-1 IF
OMUO&MUit
Ethernet
HUB
(2n)
MMDU
(n+)
IPPU
(n+)
OMU
:
O&M
Blade
•
Node internal Ethernet connectivity between functional units. External IF
(2n)
FRIU
Flexi Network Gateway highlights Market
leading capacity for Mobile Broadband growth
Market
-
leading
capacity
for
Mobile
Broadband
growth
•
ATCA-based high-performance platform
•
Up to 720 Gbps throughput & 72 M sessions per cabinet
•
Leading throughput capacity, session density, signaling performance, and DPI performance / service intelligence DPI
Flexi NG Hardware Components and Roles
The main HW components are:
ATCA ATCA ShelfShelf
•Nokia ATCA Cabinet, housing 3 shelves.
•ATCA Shelfwith 16 slots for blades. One s helf is one Flexi NG element
•Hub Blades (1+1) providing the internal switching functions
•Management Blades (1+1) running O&M and routing, and hosting disk storages
•Interface Blades providing aggregated physical connectivity
•Service Blades hosting the user sessions, running the gateway application
ATCACabinet
(CAB216)
•Service Aware Blades performing Deep Packet Inspection (DPI) functions (optional)
Service / Interface /
Service
Aware
Blade
(AMPP
2
A)
Hub Blade
(AHUB
4
-
A)
MtBld
(A
CPI CPI
5 5
A) A)
Service
Optimize the Traffic and Monetize Your Network
Compared to a standalone solution, the integrated IP Edge option, provisioning In-Line Services on a single platform, represent s the
most efficient and cost-effective way for mobile operators to optimize and monetize their networks. The integrated approach cre ates
new revenue streams, simplifies network topology and management, and lowers OPEX and CAPEX.
Integrated IP Edge
O&M
Simplifies network
topology and reduces
su
pp
ort re
q
uirements
Standalone Inline Network Topology
Several O&M O&M
S/P Gateway
Inte
g
rated
Reduce CAPEX and Improves Network
eliability under one
platform
pp q
S/P Gateway
Standalone Inline Services
NAT FW Policy &
DPI Load
g
NAT, FW,
DPI, Policy
& Charging,
Reduce
Mobile gateway needs to scale in 4 key dimensions
Th h t
Signaling capacity
Th
roug
h
pu
t
•Mobile Broadband drives throughput.
•Throughput has to be retained
independently of the signaling load
Signaling
capacity
• Flat network architecture increases
gateway signaling requirements
• Additional signaling for service
and subscriber density
awareness
• High subscriber density
• Always-on smart devices and mobility
Service intelligence • Scarce radio resources and a variety of
til i tk
b d i t lli
t
erm
ina
ls requ
ire ne
t
wor
k
-
b
ase
d
Flexi NG meets all gateway dimensions with
linear scaling by the number of the blades
Throughput
•20 Gbps per blade with Internet traffic mix (512 B
packets)
Signaling capacity
• 100 k bearer transactions per sec per shelf
packets)
•240 Gbps per shelf (element)
•720 Gbps per cabinet
• 300 k per cabinet
Platform needs to provide high performance, scalability and flexibility in all dimensions. Failure Session density
Service intelligence
Platform
needs
to
provide
high
performance,
scalability
and
flexibility
in
all
dimensions.
Failure
to perform in any of them will lead to suboptimal solution restricting operator business
severely
• 2 M PDP contexts / bearers per blade
• 24 M per shelf (element)
• 72 M per cabinet
• Low-footprint ATCA shelf (3 per cabinet)
Service
What Analysts tell about us …
Statements and endorsements
“N ki h h
“Flexi NG and Flexi
“Nokia’s mobile
packet core solution is
very threatening in the
market because it
“N
o
ki
a
h
as s
h
own
leadership in
virtualizing the EPC
with their Liquid Core
architecture. The EPC
nodes are run virtually
Nokia now a leader in the end
“Nokia has made
the product
transition and appears very
NS are based on
ATCA hardware
technology and uses
commercial-off-the-
shelf (COTS)
market
,
because
it
was one of the early
EPC solutions and
has maintained
market momentum
wi
t
h m
ajo
r m
ob
il
e
nodes
are
run
virtually
on the commercial
ATCA blades, and are
therefore adaptable to
different platforms. As
a practical matter the
leader
in
the
end
-
to-end LTE (including EPC) - Magic Quadrant
appears
very
prepared to defend
and compete ……...
It has basis of a
competitive product
tdh f
components for
deployment flexibility,
fast introduction of
technology updates
and long component
lif ti ”
tajoobe
operators based on functionality and quality of experience, and now with class leadin
g
Nokia Packet Core Leadership and Market Success
Packet Core
customers
Countries we
work in
EPC ready
product
delivered
EPC References
Global subscriber base
delivered
290
+
120
+
1
.
3
Bn
1200
+
67
Nokia has strong presence in all leading LTE markets
Innovation leadership
Industry best EPC products for LTE
290
+
120
+
1
.
3
Bn
1200
+
67
Outperforming competition
in leadin
g
LTE/EPC market
LTE
markets
.
Industry leading Telco Cloud evolution on COTS & IT h
a
r
d
w
a
r
e
Industry leading signaling capacity in EP
C
Nokia EPC in cloud
The design principles
Commercially available and operator chosen
CPU cluster (x86 farm) for cloud deployment
SGSN/MME
S/P Gateway
Nokia Cloud management for automated installation, scaling, re-use of VM and resource management in cloud environment
d Balancer
d Balancer
SGSN/MME
S/P
Nokia Core Network Nodes in the Cloud – FlexiNS
FlexiNS MME on Cloud – Virtual Machines types
Network Element
Network Element
ii
Interface blades with
elastic / scalable Load
Balancing forwarding
Stateless service
logic enables
ffi i
Open database
technology for
high capacity
FlexiNS MME vNE
Integration of vEPC
VNFs with VNFM and
Orchestrator systems
FlexiNS VM
Marker &
FlexiNS VM
Load Balancer
FlexiNS VM
Transaction
FlexiNS VM Subscriber
FlexiNS VM
O&M unit
management
stat
ist
ics
Balancing
,
forwarding
and interface
aggregation
e
ffi
c
ient,
horizontal scaling
high
capacity
database
solutions
FlexiNS VM
Network
Orchestrator
Nokia Core Network Nodes in the Cloud – FlexiNG
FlexiNG on Cloud – Virtual Machines types
Network Element management and
Intelligent IP Ede
(S/P-GW function
including policy,
Deep Packet
Logical interface
aggregation,
Database
FlexiNG vNE
FlexiNS VM
FlexiNS VM
FlexiNS VM
Interface
FlexiNS VM
FlexiNS VM
management
and
statistics
charging, SPI,
termination of all
3GPP interfaces)
Inspection
Routing and
forwarding.
Database
Instances: 2
Scalable
Scalable
Scalable
Instances: 2
Gateway Node
GWN
Service Aware
Node
SAN
Interface
Nokia Core Network Nodes in the Cloud - FlexiNG
FlexiNG on Cloud - Why to implement GW user plane in Cloud?
Because it’s possible
Mii t
Oth d it l d
... and it creates benefits
At td
E ffi i
86
M
ass
i
ve
i
mprovemen
t
s
in x86 performance:20x
improvement in Intel
packet processing in just 4
years with e.g. DPDK
Oth
ers
d
o
it
a
l
rea
d
y
today: Google and
Facebook process tens
of petabytes of data
daily already with x86
A
u
t
oma
t
e
d
deployment and
elasticity: Only possible
with pure software
based approach
E
nergy e
ffi
c
i
ency: x
86
has superior power
management over
dedicated hardware
Nokia GW is already software based: Same
software runs also on
86
hd
Gateway is not user plane limited: In most
networks, gateway is
tll liitdd
High interaction with CP
and UP: Charging,
analytics, DPI and policy
control require tight
Automatic hardware
evolution: Intel x86
improves constantly and
lHWdttii
x
86
Key Takeaway
Nokia EPC software
architecture is cloud
ready today
Dual cloud stack and COTS IT hardware support
Nokia cloud EPC solution is SDN agnostic
ready
today
Full software continuity from ATCA to Cloud
support Open approach from Nokia to provide freedom of choice to cloud deployment
Pure Software approach. No dependency on underlying SDN technology
cloud