LTE-EPC Overview - Final

EPC update
(InternaL)
4th June 2015
Agenda
› Overview of Evolved Packet Core
– Basic Architecture evolution
– Basic Concepts of PDP context & bearers
› Ericsson EPC Solution
– Impact of LTE-Advanced
– Product Overview
– Key features
– Ericsson Unique features
› Voice solution for LTE
– When to do CSFB or VoLTE?
› LTE & VoLTE Roaming
› Techno commercial discussion
› vEPC
– Cloud Intro
– vEPC
– TCO
EPC Booster | Commercial in confidence | 4th June 2015

Ericsson Packet Core
HARDWARE PLATFORMS
SGSN-MME Packet Gateway

SAPC (PCRF) DPI & Security
MSS / IMS L2/L3 PE Router
UDM Border Gateway
Evo Controller (RNC) WMG/ ePDG
Ericsson Blade System Smart Services Router
CONTROL PLANE MEDIA / DATA PLANE
”EBS” ”SSR”
500 Gbps/ 160 Gbps (DPI)
MkX: 240kTPS
~150kTPS (LTE) 30M PDNs
*capacities based on Ericsson default traffic profiles
Ericsson packet core market
share
3.6%
17.3%
Ericsson
37.7% Huawei
1.4%
Nokia Networks
ZTE
8.0%
Alcatel-Lucent
Cisco
Other
19.2%
12.8%
*Source: Infonetics (based on Annual Revenue)
http://bic.internal.ericsson.com/sources/infonetics/2015/index.php?type=reports&year=2015&service=0998

Evolved Packet Core (EPC)
architecture overview
Fewer nodes handling data plane
Separated control and user planes External
Networks
› PGW/SGW
– Payload handling HSS
› PGW: packet inspection, IP address allocation

› SGW: local mobility anchor
› MME
– Signalling node
PGW/
› Session management MME SGW PCRF
› Security
› Bearer control
› HSS
– Subscription data storage and management
LTE
› PCRF
– Policy controller
High performance differentiated service offering over LTE

Ericsson Packet Core evolution
Mobile Broadband Boosting (2G/3G) Superior Experience with EPC/LTE
IP IP
Networks Networks
SAPC SAPC
EPG
Charging GGSN Charging
SGSN
Introduce 4G access SGSN
WiFi
Fixed access
HSS MME
HLR Pool
Pool
WiFi CUDB
Fixed access
LTE
GSM GSM
WCDMA
WCDMA
› Service Aware e2e QoS – (Packet Inspection & PCRF)

› Signalling Optimizations to handle Smartphones › LTE introduction via Common EPC
› Network Integrated WiFi › Video Optimization - CDN/TIC
› SGSN Pooling for network robustness › Voice and MMTel support
› IPv6 deployments
Enabling a smooth migration to LTE/EPC

2G/3G MIGRATION TO EPC
Migration scenarios – FULL Upgrade Proposed when we are incumbent
› Common core for optimal EPC capacity usage across IP

Networks
access technologies
› GGSNs configured as GGSN/PGW/SGW PCRF
› SGSNs configured as SGSN-MMEs DB
› PCRF providing support for the required QoS GW

GW
mechanisms HLR-FE HSS-FE
Key benefits
› Smooth network upgrade
› Common Packet Core
SGSN MME
› Minimal changes to existing charging, O&M and policy
control systems
› Optimum capacity usage
2G 3G LTE
› Reduced signaling load
› Dynamic capacity sharing reducing CAPEX 2G
3G
2G
3G
LTE

2G/3G MIGRATION TO EPC
Migration scenarios – overlay Proposed to break-in
› Existing SGSNs and GGSNs kept for legacy IP

Networks
GSM/WCDMA terminals
PCRF
› New MME and Packet Gateway serve LTE-capable
terminals DB
GW
GW
› PCRF providing support for the required QoS GGSN
HLR-FE HSS-FE
mechanisms
› Interworking between GSM/WCDMA and LTE
› Multi-mode LTE-capable terminals
can connect over GSM/WCDMA SGSN MME
when no LTE is present
› Different GW selection methods
2G 3G LTE
Key benefits
2G
› Low risk – existing network not affected 3G
› Gradual transition 2G
3G
LTE

Network topology Might evolve from 1 scenario to the other
Centralized SGSN/MME/GGSN/GW Distributed SGSN/MME, Centralized GGSN/GW
1 2
Pros: Pros:
 Less network elements  SGSN-MME with pool => Improve the capacity,
 SGSN-MME with pool with fewer required nodes reduce signaling, increase redundancy
 Good for small network operator/vertical segment  Good for medium scale operator
Cons: Cons:
 No Geo-red  No Geo-red for GW
 High transport cost  High transport cost on user plane
Distributed SGSN/MME/GGSN/GW Distributed GGSN/GW, Centralized SGSN/MME

Pros:
3 4
Pros:
 SGSN-MME with pool => Improve the capacity,  SGSN-MME with pool with fewer required nodes
reduce signaling, increase redundancy  Efficient cost on transport
 Efficient cost on transport  Good for medium scale operator
 Full Geo-red Cons:
 Good for big operator  High cost on 2G transport
Cons:
 High Capex on NW elements

Network Topology
general observations
› Centralization of the MMEs is a good strategy in terms of reducing the operational costs
– Centralized MMEs in "server farms" simplifies O&M, reduce costs both due to reduced no of sites and less
distribution of operational knowledge
› Centralized SGW/PGWs is suitable for an introduction of LTE according with limited coverage and/or
limited subscriber penetration
– Synergies with existing Mobile networks topology, allowing for re-use of sites, transmission, peering points,
connections to services etc
› As traffic volumes grow, thresholds for when it is beneficial to first partial and then full distribution of the
SGW/PGWs in the network instead of having them centralized may be passed. Factors that influence:
– Service and user-related parameters, e.g. Population density, Offered services and assumed application mix, Traffic
model assumptions etc.
– Network-related parameters, e.g. Number of existing and available sites, site distribution, existing Internet peering
points etc.
› Distibuted GWs is also suitable for converged service operators especially when using LTE as DSL
complement or replacement
– Synergies with existing Fixed networks topology, allowing for re-use of sites, transmission, peering points,
connections to services etc
Might be different case to case, depending on location/no. of PoPs, existing node locations,
multi-vendor, organizational requirements
Network topology
split or combineD S/P-gateways
Distributed SGW & PGW
Combined SGW & PGW
PDN
GW PDN
GW
SGW SGW SGW SGW
MME MME
MME MME
MME
MME
› Implemented by few operators due to: organizational responsibility, › Implemented by most operators
preference to source from different vendors
› All traffic pass through both SGW and PGW in central location
› Most traffic will be routed back to centralized PGW
– Roaming traffic will traverse S8 to Home PGW › For LTE roaming case, nodes need to support S8 and work as
› Distributed PGW may be implemented towards existing Internet standalone SGW and PGW
peering points etc. – physical split is not required
› SGW located close to eNBs, acting as local mobility anchor when › Lower TCO (fewer nodes, SW upgrades, floor space, regression
users are moving between eNBs
tests etc)
› Higher TCO (more nodes, more O&M etc)

What’s new in EPS QoS
› Network-centric bearer control
– Network initiated Dedicated Bearers
– Uplink packet filters installed by the network into the terminal
› No QoS negotiation between E-UTRAN and EPC

– MME can reject, but not re-negotiate QoS
› New QoS Parameters

– Class-based mapping of operator services to packet forwarding treatment,
› Introduction of 9 standardized QCI (QoS Class Identifier)
– Aggregated MBR
› APN-AMBR - Aggregated MBR per APN
› UE-AMBR - Aggregated MBR per UE

Bearer QoS Parameters:
QoS Class Identifier (QCI)
› 9 Standardized QoS Classes
› In Specs Each of the 9 QCIs is Associated with Standardized Characteristics in Terms of Priority, Packet
Delay Budget, and Packet Error Loss Rate
› Operator Policy Decides Semi-Static Mapping Service+Subscriber-ID to QCI
QCI Resource Type Priority Packet Delay Budget Packet Loss Rate Example Services
1 GBR 2 100 ms 10-2 Conversational Voice

2 4 150 ms 10-3 Conversational Video (Live Streaming)
3 3 50 ms 10-3 Real Time Gaming
4 5 300 ms 10-6 Non-Conversational Video (Buffered Streaming)
5 Non-GBR 1 100 ms 10-6 IMS signaling

6 6 300 ms 10-6 Video (Buffered Streaming)
TCP-based (for example, www, e-mail, chat, ftp, p2p
file sharing, progressive video)
7 7 100 ms 10-3 Voice, Video (Live Streaming), Interactive Gaming
8 8 300 ms 10-6 Video (Buffered Streaming), TCP-based (for example,

www, e-mail, chat, ftp, p2p file sharing, progressive
9 9 video)

Basic concepts
sessions, pdp context and bearers
IP Session
Default Bearer / Primary PDP Context
Service 1, e.g. Internet browsing

Service 2, e.g. P2P File Sharing
Service 3, e.g. Streaming
Terminal Core
Equipment RAN Transport Network Dedicated Bearer / Secondary PDP Context
User SESSION: created in each PC node. ONE APN = ONE Session = one IP address
One primary PDP to provide basic connectivity One default bearer to provide basic connectivity
One or more secondary PDP to provide different QoS One or more Dedicated Bearers to provide different
QoS
Support on 3G terminals started, on Qualcomm
chipsets Available on LTE-capable terminals from Day 1

Basic concepts
charging mechanism (software)
IP Session
Default Bearer / Primary PDP Context
Service 1, e.g. Internet browsing

Service 2, e.g. P2P File Sharing
Service 3, e.g. Streaming
Terminal Core
Equipment RAN Transport Network Dedicated Bearer / Secondary PDP Context
SGSN: kSAU, IP session, kpps MME: kSAU, IP session

EPG: IP Session, kpps/Mbps EPG: IP Session, kpps/Mbps
IP session: Only Primary PDP contexts are IP Session: Only Default Bearers are
counted for SW licensing purpose counted for SW licensing purpose
(Secondary PDP contexts are not (Dedicated Bearers are not considered).
considered).

Agenda
– Key features
› vEPC
– Cloud Intro
– vEPC
– TCO

Epc aspects – lte advanced
› Traditionally, Packet Core is normally never a
performance bottleneck (if properly dimensioned)
› ...but this may not always be the case in the future

– New LTE devices with 300-450 Mbit/s peak throughput hitting
the market
New & more Increasing challenges

spectrum
 Capacity
More advanced
devices & services  Peak rates
 Latencies
More & smaller cells

3G/4G technoLogy evolution
› Data rate evolution
– Higher mean data rates HSPA
– Higher peak rates
› How to serve these data rates in an efficient way?

– With and without DPI
CAT6 the challenge in 2014
LTE LTE
CAT4 the challenge in
2011/ 2012
LTE Advanced
Cat-6 devices
(300/50 Mbps)

Considerations when implementing
LTE/EPC
› Increased usage of apps and real-time services (device and network support it!) e.g. Voice, Video
Higher throughput
Higher signaling
More active bearers (Dedicated bearer support)
› Greater need for operator to get better business intelligence & faster troubleshooting
Event Based Monitoring & reporting
Packet Inspection & heuristics
› More critical to ensure: How to address
High availability/Network performance
E2E QoS these
Seamless mobility
› Voice calls - using CSFB/VoLTE/SRVCC/WiFi calling means even more interworking with
CS & IMS ecosystems
› LTE Roaming

Ericsson SGSN-MME
High level capacity evolution overview – Signaling (LTE TPS)
MkX
(6th generation HW)
GA: July 2015
Smallest footprint per subscriber
GEP5
GEP5-SSD
SMXB
MkX
(target)
MkVIII
Major leap
MkVIII
MkVIII
MkVI ~150kTPS
MkVI MkVI (LTE)
13A 13B 15A

EPG on SSR
today >200 contracted/deployed
EPG EPG EPG EPG
2010 2013/14 2015/16 Future
Chassis/backplane
capacity with new
blade versions
1x 10x 20x 400x

EPC Booster | Commercial in confidence | 4th June 2015 Throughput figures are Full Duplex and with Ericsson traffic model
SSR ARCHITECTURAL VERSATILITY
FLEXIBLE SCALING OF RESOURCES
SCALES FOR SIGNALING & FASTER FEATURE VELOCITY VIRTUALIZING APPLICATION

FORWARDING AND APPLICATION ENVIRONMENT
INTEGRATION
THROUGHPUT MIX OF FORWARDING AND APPLICATION APPLICATION

CENTRIC SYSTEM CONTROL PLANES CENTRIC SYSTEM MODULE ON VMs
ePDG Wifi GW
SASN MSP/CD
Wifi
ePDG GW/ePDG (DPI) N
EPG 3rd partiy

SASN MSP
EPG BNG
BNG EPG
Available: EPG & router, EPG & ePDG

This year: EPG & WMG, EPG & MSP, EPG & SASN
=> For the combinations beyond EPG + Site Router contact EPG product management before offering

Comparison to other vendors
Huawei NSN ALU Cisco ZTE
High power Flexi-NS (ATCA): Limited 2G/3G feature Limited references Poor product quality; few
SGSN-MME
consumption & large Does not support 2G/3G set, more suitable for references;
footprint 4G overlay reqs 2x power and
For 2G/3G, need older footprint vs ///
MkVIII vs OSTA 2.0 (9.x) platform DX-200
For 30M bearers, we are
33% of Huawei’s
footprint
High capacity reduction Flexi-NG (ATCA): Limited 2G/3G feature Different platforms for Poor product quality; few
with DPI Dedicated blades for set, more suitable for router & gateway references
Gateway
DPI 4G overlay
Per blade: 5Gbps@512
B with 100% DPI 50% capacity reduction
EPG 14B: 9 Gbps@650 with DPI
B with 100% DPI
Uncertainty due to Nokia/ALU integration

Key Benefits of Triple
Access Pooling
SGSN/MME Pool
› Reduced signaling load and High Availability SAU SAU SAU
• Good user experience SGSN SGSN SGSN

MME MME MME
› Same hardware and software for all Access

• No feature disparity
BSC BSC RNC BSC
• Flexibility in subscriber migration
• Maximum reuse of existing investment for RNC eNB eNB eNB
future growth
Commute effect SAU in the pool members
SAU
› Minimal changes to existing charging, O&M and

without Pool
policy control systems
• Faster time to market SAU
#1 #2 #3 #4 #5 SGSN-MME
› Ericsson proprietary subscriber move with with Pool traffic shared among SGSN-
MMEs
session and traffic continuity mechanism. #1 #2 #3 #4 #5 SGSN-MME
traffic in day-time traffic in night-time
EPC Booster | Commercial in confidence | 4th June 2015 Same UE used for both Vendors
Reference Case Study
Common core VS overlay Accum ulated CAPEX
3,500,000,000
3,000,000,000
2,500,000,000
2,000,000,000 Common C ore

Overlay
1,500,000,000
1,000,000,000
500,000,000
-
2013 2014 2015 2016 2017 2018
Accum ulated OPEX
1,200,000,000
1,000,000,000
› Triple-access provides 20-30% CAPEX/OPEX 800,000,000
savings Common Core
600,000,000
Overlay
› both
Dynamichigher CAPEX
and automatic &forOPEX
scaling optimized use of 400,000,000
capacity 200,000,000
-
2013 2014 2015 2016 2017 2018

TCO Assumptions

Signaling Optimization & Control
signaling reduction & control features (For LTE/EPC)
Dynamic TAI list Smart and Adaptive UE Signaling Control

per UE paging
› Reduced paging and › Up to 85% less paging › Protection against

TAU procedures with › Increasing the active users node overload situation
typically 25% in SGSN-MME with up to and devices with
100% abnormal traffic
patterns
ERICSSON PATENTED FUNCTIONALITY

Signaling Optimization & Control
customer cases
Excess Messages Threshold 30 during 60 s
Smart and V * eNB UE Signaling 400000
Intra MME TAU
300000
Adaptive paging Control 200000

Service Request
Attach Request
X * eNB list 100000
PDN Connectivity
0 Request
2013-06-07
2013-06-10
2013-06-13
2013-06-29
2013-07-02
2013-07-05
2013-06-01
2013-06-02
2013-06-03
2013-06-04
2013-06-05
2013-06-06
2013-06-08
2013-06-09
2013-06-11
2013-06-12
2013-06-14
2013-06-15
2013-06-16
2013-06-17
2013-06-18
2013-06-19
2013-06-20
2013-06-21
2013-06-22
2013-06-23
2013-06-24
2013-06-25
2013-06-26
2013-06-27
2013-06-28
2013-06-30
2013-07-01
2013-07-03
2013-07-04
2013-07-06
2013-07-07
2013-07-08
Y * TA
case: Z * TA list case:

Softbank japan North American operator
Other UE by Vendor
5%
FUNCTION ACTIVATION
Unknown Samsung
26% 34%
Apple
Huawei 14%
2%
LG
7% HTC ZTE Toshiba
6% 1% 5%
 75% paging reduction leads to an improved
MME capacity by 31%  Devices generating 10 times more
signaling than average
 Based on subscriber forecast a prolonged
investment by 14 months !!  Discovered and network protected in 3-4
minutes vs days

 Supports UE lockout
Mobility based policies
› Direct communication between the SGSN-MME and the SAPC - Sx i/f
› SAPC enriches policy decisions with the latest user location information and SRA (Sx – Presence Reporting Area) status
of the user sent by SGSN-MME.
› SAPC is able to select the preferred GW, or RFSP, based also on ToD, device type or subscriber data usage, for better
network resource efficiency.
› Use cases: Location based use cases: Home-Zone Broadband, Mobile Broadband service for Enterprises, Rural areas
usage promotion, balance radio usage in multi-LTE Policy Decision
Mobility Management
Mobility-based Policy
SGSN-
MME SAPC
* SGSN-MME 16A for GSM/ WCDMA Radio

Access GW
30% Reduced load

Geo-redundant pool
towards 100% Service availability
GW
S11
User data Virtually no service impact on end users

is replicated
even in outage scenarios
MME MME Replication of user data & status:
- from MME to MME
S1 - from Sub-Pool to Sub-Pool
Designed to have minor impact:

CPU (below 10%)
Bandwidth (<70 Mbit/s at 3.7 MSAU)
eNB
EPG supports MME Geographically Redundant pool function -
immediately selects alternate MME at MME failover
service continuity
Packet Handover Restoration Procedures
3G LTE M2M IMS
 Minimizes customer impact

0,2 sec at network problems.
GW GW  Secure end-user experience
No service impact
and secures revenue.
 Maintains or re-activates
 No end user impact when moving. connection:
 Enables streaming services e.g NetFlix,
›At link problems
Skype and Vimeo
 Is needed to keep customer satisfaction. MME ›At GW problems
 Requires support on SGSN-
MME & EPG

Event based monitoring
overview Service Assurance
Network Intelligence
Service
Business Intelligence
Assurance
Event Based Monitoring
(EBM): real-time monitoring of
ALL users ALL the time SGSN-
MME SAPC
›Streamed data feeds

›Limit end user impact by instant action eNodeB
RNC
EPG SASN Internet
at any performance issue

›Guided fault isolation and resolution by › >100 operators live with EBM
Ericsson injected know-how › 15% faster customer issue resolution
›Multi-dimensional / e2e correlations › Minimized Subscriber Churn
provide full end-user insights for
business intelligence purposes › 75% lower CAPEX/OPEX than probe solution
› Data feeds made available to any Service Assurance
application with monitoring and reporting capabilities

service
distribution
per terminal
type
url
distribution
per ue
location
Agenda
– Key features
› vEPC
– Cloud Intro
– vEPC
– TCO

Epc support for volte
› EPC is a central part of the VoLTE
solution
– Service announcements
– IMS node selection Sv MSS
– Control of dedicated bearers S6a

– SRVCC HSS
– Emergency calls
EPC
– Robustness & resilience
MTAS
Rx MRFC
NAS
IMS
MRFP
PCSCF
QCI=5 SIP Signalling Gm
QCI=1 Voice Mb
QCI=2 Video AGW
IMS APN

VoLTE trends 3GPP
Evolution steps in a telephony perspective
Spotty LTE coverage Areas with continuous Continuous IMS voice over PS
LTE coverage coverage
LTE
CSFB LTE LTE
LTE LTE LTE
LTE LTE LTE LTE
LTE LTE LTE LTE LTE LTE
LTE LTE LTE
LTE LTE LTE
LTE
2G/3G 2G/3G LTE

2G/3G LTE
LTE
SRVCC
LTE
LTE
LTE
3G / HSPA
LTE
LTE
LTE
LTE
LTE
LTE
LTE
LTE
LTE
LTE
LTE
LTE
LTE
LTE LTE
LTE LTE LTE LTE LTE
LTE LTE LTE LTE LTE LTE
LTE LTE LTE LTE LTE
PS voice PS voice
Legacy
Phone CSFB PS voice VoLTE
CSFB VoLTE SRVCC LTE
CS voice MSS Phone CS voice MSS Phone
SRVCC IMS Phone IMS
CS voice & ICS HSPA
MSS
IRAT PS HO
CS voice service via CSFB IMS Voice service over LTE and IMS Voice service over LTE
CS access using SRVCC and HSPA
CSFB = Circuit Switched Fallback

SRVCC = Single Radio Voice Call Continuity
ICS = IMS centralized services

When to Use What
High probability of losing LTE
coverage during a call
CSFB
CS Telephony
SR-VCC
IMS Telephony
with PS to CS
handover
Very low probability of losing IMS Telephony

LTE coverage during a call
Without SRVCC
› Important factors:
– Use of frequencies for LTE / Quality of coverage
– Underlying CS and PS network

CS Fallback
• Data Service
› Uses existing Circuit-Switched voice • Voice call
• Data Service
services (GSM, WCDMA)
– Includes emergency services
– No IMS system is needed
› Automatic transfer of LTE user to CS at PDN
incoming/outgoing call
CS/PS EPC
› Also parallel data sessions may be
transferred CS Fallback
– Continued high-speed, seamless service transfer 2G/3G or
using Packet Handover from LTE to HSPA LTE
CDMA
› Requires SGs interface between MME and
MSC

Pooling & CS Fallback
- the ideal combination
› Supports MME Pooling, SGSN Pooling and PDN
MSC in Pools – all combinations

– The MMEs always detect available MSCs for secure
voice call establishment MME
MSC
› Full service availability using Pooling MSC MSC MME MME
CS Fallback
– ISSU: Maintenance without voice service interruption
– Geo-redundancy with voice service continuity also at
SGSN
unplanned events such as natural disasters SGSN SGSN
– 20-30% reduced network signaling – compensates for

the additional CS Fallback mobility signaling
– 20-50% node capacity savings with Uniform load
balancing

SRVCC - secures full voice coverage
irrespecive of access technology
• CS call • VoLTE call
• Data session
› Continued voice service when user moves out • Data session
of LTE coverage CS PDN

IMS/
PDN
– Automatic transfer from VoLTE to CS
– Includes emergency services
› Also parallel data sessions may be transfered CS/PS EPC
SRVCC
LTE 2G/3G or
LTE
LTE LTE CDMA
LTE
LTE
SRVCC

SMS over SGs
Concept
› SMS in LTE is the same service as in GSM and WCDMA

– Content limited to what SMS services provides
› n x 160 characters, EMS (small pictures, sounds, animations and formatted text)
› The same delivery architecture is used

– The SMS is uploaded to the SMS-SC on the originating side
– The SMS is delivered to the delivery point based on an HLR interrogation
Originating Terminating
HLR GSM
MAP MAP
DTAP RANAP/NAS WCDMA
MSC SMS-SC MSC

LTE
› New interface towards the EPS domain (SGs-interface) MME
– Combined CS/EPS registrations for LTE users with CS based services

– Delivery of SMS’s between EPC and MSC (both MT and MT SMS’s)

Epc support for volte
› EPC is a central part of the VoLTE
solution
– Service announcements
– IMS node selection Sv MSS
– Control of dedicated bearers S6a

– SRVCC HSS
– Emergency calls
EPC
– Robustness & resilience
MTAS
Rx MRFC
NAS
IMS
SGSN-MME PCRF
MRFP
PCSCF
QCI=5 SIP Signalling Gm
QCI=1 Voice Mb
QCI=2 Video AGW
EPG IMS APN

Agenda
– Key features
› vEPC
– Cloud Intro
– vEPC
– TCO

LTE Roaming Architecture
Home Tunneling vs Local Break Out
Home Tunneling Local Break Out
IP IP
Networks Networks
H-PCRF H-PCRF
DB DB
PGW PGW S9 (optional)

HSS HSS
S8
HPLMN S6a HPLMN S6a
VPLMN VPLMN
MME SGW MME SGW V-PGW

V-PCRF V-PCRF
LTE LTE

Lte roaming
usage of a diameter EDGE Agent (DEA)
› LTE roaming relies on inter-operator Diameter connections between EPCs in

HPLMN and VPLMN
– S6a, S6d, S9
› Roaming Edge Agent part of GSMA PRD IR.88 Supported with Ericsson DSC
– Scalability
– Resilience
– Topology hiding
VPLMN HPLMN
MME S6a S6a HSS

S6d
S6d
SGSN DEA DEA
GRX/IPX
S9 S9
vPCRF hPCRF

Agenda
– Key features
› vEPC
– Cloud Intro
– vEPC
– TCO

Techno-Commercial
Competitor measures capacity in different way
› Vendors are having different interpretation of how to Failed Connections

measure traffic › One IP session is charged even if the call-setup failed
› Most RFQ:s are not considering What, Where, and › Could mean 50% more IP sessions charged in markets
How to measure capacity with high penetration of China and India produced terminals
– Only providing number of nodes and traffic projections
IP
Incoming and Outgoing Traffic Networks
› Typically used when the interfaces are not specified SGSN-MME EPG
and is used for CS and PS Core to double the charged

capacity Signaling traffic
› If not specified, some vendors are including non-payload
IP
Networks signaling traffic into the Mbps/PPS measurements
UE is SGSN-MME EPG
sending › Effect could be up to 30% depending on terminal type and
and Charged by network topology
receiving Ericsson
data
IP
Charged in addition Networks
by other vendors UE is sending and SGSN-MME EPG
receiving data
Techno-Commercial
Competitor measures capacity in different way
› Some vendors are SoC compliant in providing triple-access nodes

– When implementing the HW they however claim that HW is slightly unstable and offer to give HW
nodes FOC [HW/SW Price Split: ~20/80]
› Having separate nodes for GSM/WCDMA/LTE traffic means up to 3 times more licenses having
to be provisioned for in order to support traffic pushed/falling back between different
technologies
› Even when supporting triple access in a single node, they are charging SW licenses for:
– SAU - separately for each access type
– PDP/ Bearers - separately for each access type 2G
› In addition, they are also charging for dedicated bearers (& secondary PDP) 3G
SGSN-MME
LTE /EPG
Triple access

Techno-Commercial
comparison between vendors
Major influence
Use RASO Developed Normalization Kit in customer discussions

Link to value calculators
http://146.250.214.254/ValueCalculator/login_value_calculator.php
Register first!

Agenda
– Key features
› vEPC
– Cloud Intro
– vEPC
– TCO

LTE-EPC Overview - Final

Uploaded by

Copyright:

Available Formats

LTE-EPC Overview - Final

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

LTE-EPC Overview - Final

Uploaded by

Copyright:

Available Formats

EPC update

EPC Booster | Commercial in confidence | 4th June 2015

SGSN-MME Packet Gateway

Ericsson Blade System Smart Services Router

CONTROL PLANE MEDIA / DATA PLANE

*Source: Infonetics (based on Annual Revenue)

EPC Booster | Commercial in confidence | 4th June 2015

› PGW: packet inspection, IP address allocation

High performance differentiated service offering over LTE

› Service Aware e2e QoS – (Packet Inspection & PCRF)

Enabling a smooth migration to LTE/EPC

› Common core for optimal EPC capacity usage across IP

› SGSNs configured as SGSN-MMEs DB

› PCRF providing support for the required QoS GW

EPC Booster | Commercial in confidence | 4th June 2015

› Existing SGSNs and GGSNs kept for legacy IP

EPC Booster | Commercial in confidence | 4th June 2015

Distributed SGSN/MME/GGSN/GW Distributed GGSN/GW, Centralized SGSN/MME

EPC Booster | Commercial in confidence | 4th June 2015

SGW SGW SGW SGW

EPC Booster | Commercial in confidence | 4th June 2015

› No QoS negotiation between E-UTRAN and EPC

› New QoS Parameters

EPC Booster | Commercial in confidence | 4th June 2015

1 GBR 2 100 ms 10-2 Conversational Voice

5 Non-GBR 1 100 ms 10-6 IMS signaling

7 7 100 ms 10-3 Voice, Video (Live Streaming), Interactive Gaming

8 8 300 ms 10-6 Video (Buffered Streaming), TCP-based (for example,

EPC Booster | Commercial in confidence | 4th June 2015

Service 1, e.g. Internet browsing

EPC Booster | Commercial in confidence | 4th June 2015

Service 1, e.g. Internet browsing

SGSN: kSAU, IP session, kpps MME: kSAU, IP session

EPC Booster | Commercial in confidence | 4th June 2015

EPC Booster | Commercial in confidence | 4th June 2015

› ...but this may not always be the case in the future

New & more Increasing challenges

EPC Booster | Commercial in confidence | 4th June 2015

› How to serve these data rates in an efficient way?

CAT6 the challenge in 2014

EPC Booster | Commercial in confidence | 4th June 2015

EPC Booster | Commercial in confidence | 4th June 2015

EPC Booster | Commercial in confidence | 4th June 2015

1x 10x 20x 400x

SCALES FOR SIGNALING & FASTER FEATURE VELOCITY VIRTUALIZING APPLICATION

THROUGHPUT MIX OF FORWARDING AND APPLICATION APPLICATION

EPG 3rd partiy

Available: EPG & router, EPG & ePDG

EPC Booster | Commercial in confidence | 4th June 2015

Uncertainty due to Nokia/ALU integration

EPC Booster | Commercial in confidence | 4th June 2015

› Reduced signaling load and High Availability SAU SAU SAU

• Good user experience SGSN SGSN SGSN

› Same hardware and software for all Access

› Minimal changes to existing charging, O&M and

2,000,000,000 Common C ore

Accum ulated OPEX

EPC Booster | Commercial in confidence | 4th June 2015

EPC Booster | Commercial in confidence | 4th June 2015

Dynamic TAI list Smart and Adaptive UE Signaling Control