Self Organizing Networks - SOCRATES workshop Santander... · Drivers for Self Organizing Networks ... LTE RAN EM: SAE GW SQM BTS Site ... commissioning SAE-GW / PDN-GW MME Network

1 © Nokia Siemens Networks

For internal use only

Self Organizing NetworksA Manufacturers View

ICT Mobile SummitSantander, SpainJune 2009

Werner Mohr

2 © Nokia Siemens Networks For internal use only 2009-06-09SON - A Manufacturers View Dr. Werner Mohr

SON – What it is about

• SON – Drivers • LTE OAM Architecture• SON Functional Domains• Manual vs. SON-based Network Operation• Implementation Approaches


Definition

NSN’s position• Configure & optimize network automatically• Allow operators to be final control instance• SON goes beyond LTE• SON starts a paradigm shift how to

operate a network

“Self Organizing Networks” - SON


Configure & optimize the network automaticallyBut allow the operator to be the final control instance

• Configure & optimize the network automatically• But allow the operator to be the final control instance

Drivers for Self Organizing Networks

• Saturated markets, revenue per bit is dropping • Parallel operation of LTE with 2G and 3G networks• Large and complex number and structure of

network parameters• Expanding number of Base Stations

OPEX reduction – reduce human interaction• OPEX reduction – reduce human interaction


Drivers for Self Organizing Networks

fasterrollout & netw.

extensions

simplify network

operation

minimizeeffort foraddl. LTE operation

reduceOPEX

adopt tosubscriberbehavior

sustainednetwork quality

SON drivers


LTE / SAE: Flat OAM Architecture

OAM Network ManagementOptimizer MQA App nTraffica Itf-N

IntegratedOperation & Mediation

Function

BTS Site Manager EM: MMEEM:

LTE RANEM: SAE

GW

SQM

BTS SiteManager

Local connected

LMT

LMT

LMT

LMT

MME SAE-GWeNB eNB

XMLover HTTPS

BTSOMASN.1

& HTTPSNE3S/SNMP


SON Functional Domains

Self-configuration: automated network integration of new eNBby auto connection and auto configuration, core connectivity (S1) and automated neighbor site configuration (X2)

Self-optimization: auto-tune the network with the help of UE and eNB measurements on local eNB level and/or network management level

Self-healing: automatic detection, root cause analysis, localization and removal of failures

Self-planning: dynamic re-computationof network plan, e.g. due to capacityextensions, traffic monitoring oroptimization results

Healing

Optimization

Configuration

Self-Plan

ning


SON Functional Domains

Self-Healing

Self-Optimization

Self-Configuration

Plug and Play Auto Setup, -Config., Phy. Cell-ID Auto Connectivity X2, S1 Auto Neighbor Relation, ANR

Auto RepairHW/SW-Failure MitigationCell Outage Detection and Outage MitigationAutomatic Alarm Reaction

Auto TuneCoverage & Capacity, Energy savings, Mobility Robustness, Load Balancing, RACH, Inter-Cell Interference Coordination

Self-Planning

“Dynamic re-computation”due to capacity extensions, traffic monitoring or optimizations


Current Network Operation (simplified)

Rep

ortin

g Th

resh

olde

r& P

rofil

er

Mon

itorin

g

Opt

imiz

atio

n

NE

Con

figur

atio

n

External RadioNetwork Planning

CM

OAM

Serv

ices

(Too

ls)

LTE RA Access & Core Network

SW M

anag

emen

tLi

cens

e M

gmt.

PM FMCM


Rep

ortin

g Th

resh

olde

r& P

rofil

er

Mon

itorin

g

Opt

imiz

atio

n

NE

Con

figur

atio

n


OAM

Serv

ices

(Too

ls)

Workflow:eNB Auto

Configurator

Workflow:Cell Outage

Compensation

Workflow:Alarm triggered

actions

LTE RA Access & Core Network

SW M

anag

emen

tLi

cens

e M

gmt

PM FMCM

Policy machine (SON Coordinator)

Workflow Engine(Business Process Execution Language)

SON Coordinator executes pre-defined workflow-objectsorchestrating the NetAct fragments

SON enhanced Network Operation


SON Implementation Approaches

faster timescale of operation

number of cells involved

Centralized

• > 2 cells involved• Slower update rate• Long term statistics

Distributed

• ~ 2 cells involved• > 2 cells if changed

parameters have only local impact(only neighbor inforequired)for > 2 cells master/slave relations required toallow convergence

• utilize X2 signaling

Localized

• single cell scope• fast update rate• short term statistics

Crit

eria

Exam

ples

• Auto Configuration with dynamic selfplanning

• coverage and capacity optimization

• Automated NeighborRelation ANR

• HO & InterferenceOptimization

• Cell outage compensation

• adaptive MIMO• Link adaptation• RACH optimization


Nokia Siemens Networks SON Architecture

Nokia Siemens Networks goes for Hybrid SON, parts of the self-organizing functions and algorithms are executed on OAM level while others are done in eNB:

• Quick short term and simple optimization schemes in eNB

• Complex, mid & long term optimization schemes in OAM (e.g. Optimization tools)

• Flexible to support different optimization cases

• Multivendor support via X2 Itf.*)and Domain Manager northbound Interface*)

*) for standardized SON functions/Interfaces

Domain / Element

Management

Network Management

Itf.-N

X2 X2

SON SON

SON

SON

SON

DMother

Vendor

OAM


SON - Examples

• Initial eNB Configuration• Automated NW Planning• ANR Configuration• Self-healing


Example: Initial eNB Configuration

1-Touch-Deployment:• eNB default pre-config. in factory• Shipping to site• Site preparation and eNB installation by

field service – plug & play:• eNB gets automatically connected

and configured from OAM System• eNB gets automatically integrated into

the access-, core- and network-management networks

V

Factory Assembly On-site installation(Field service)

Auto connection &configuration, remotecommissioning

SAE-GW / PDN-GW

MME

NetworkManagement

eNB


OAM

SONCoordinator

AutoConnServices

Example: Initial eNB Configuration Steps

IP BackhaulIP Backhaul

1. auto-connectto backhaul

2. IP address allocation

3. Authentication & PKI Certificate

Download

4. Auto SW &

Configuration download

5. Auto X2 set-up

6. Auto S1 set-up

Certification &RegistrationAuthority

DHCPServerMME

GW


Example: Automated Neighbor Relation (ANR) configuration

• Relations between neighbor cell need to be carefully known respectively planned

• Wrong configurations cause HO failures / dropped calls• Self configuration of relations avoids manual planning &

maintenanceANR covers 4 steps:• Neighbor cell discovery• Neighbor Site’s X2 transport configuration discovery• X2 Connection Set-up with neighbor cell configuration update• ANR Optimization


SONCoordinator

Example: Automated Network planning

Planning tool delivers very few parameters, all other parameters areautonomously derived by the system, such as physical cell identifier, pilot specific settings, neighbors and other cell specific parameters by tight interworking of Configuration Management & Network Optimization

Optimization

OAMCM


Templates

Configuration

eNB Config

~5 to 10 parameters

~ 400 parameters

Polices


Example: Self Healing topics

NetAct• Self Healing is closely related to fault management, performance management,

redundancy & resilience concepts• Nokia Siemens Networks applies self healing capabilities on LTE eNodeBs,

higher level network elements and as part of NMS applications respectively on service level

LTE Base Stations: • Redundant Architecture for

– System Modules– RF Modules and RF

Heads– Power supply Module– Temperature control

(fans)• Integrated System

Maintenance for permanent supervision and diagnostic

• Responsible for alarming and proper recovery measures like SW fall back

• Evolved Packet Core pooling concept for Mobility Management Entities and Gateways

• Redundant and self routing Carrier Ethernet Transport and IP aggregation network

MME Pool

Automatedroot causeAnalysis

FM PM

Mobile QualityAnalyzer

measures

Network Management: • Root Cause Analysis

Automation• Sleeping Cell detection• Cell Outage compensation

(e.g. by Remote Antenna Tilt adjustments of neighbors)

…


Example: Cell Outage Detection / Compensation

Detection algorithm will use thresholds for raising alarms which indicate a possible cell problem

• Thresholds based on typical network traffic of a cell• Used tooling: Performance Management

Compensation by re-configurations of neighbor relations, changing interference related parameters, transmission power settings, antenna tilt setting, ...

• Used tooling: Network Optimization

Study topics:• Trigger points: when the procedure shall start • Algorithm for automatic compensation: which perspective is the

most important and which restrictions may be accepted for limited time


Wrap-up

• Time-to-market• Summary


SON Operator Benefit and Time to Market

ImpexField

Installation

ImpexPlanning

Capacity based

optimizationCoverage

based optimization

Quality based optimization

Opex ReductionNOC

Opex Reduction Field Maintenance

Ben

efit

for O

pera

tor

Low

High

Time to MarketImmediate atLTE Rollout

Later

Opex Power Savings


Summary

• Major drivers for SON are• Increased number of Base Stations in different layers (macro,

micro, pico, femto cells)• Increased number of network parameters to be taken into account• Increased network planning and management effort needed• Strong requirements on OPEX reduction

• Quality of Service and user experience will only be maintained in such complex environment by means of

• self-optimized and• self-managed networks

• SON concepts help to enable plug & play network extension and configuration

Documents

Self Organizing Networks - SOCRATES workshop Santander... · Drivers for Self Organizing Networks ... LTE RAN EM: SAE GW SQM BTS Site ... commissioning SAE-GW / PDN-GW MME Network