ALU RF Optim Process

Training session 3 Optimization services

August 2010

)

Alcatel-Lucent W-CDMA Description All rights reserved © 2009, Alcatel-Lucent

Agenda (Radio/optimization operational services) 2

Global Optimization Process Global view Mandatory task for new 3G operators

Pre-optimisation (design phase) Pre-Launch Optimization Process General view (process/KPI/ressource) Configuration Audit Site Audit Cluster Optimization Network Optimization

Post-Launch Optimization Process Optimization Tools

Workshop OCI – Services Optimisation – December 2009

All Rights Reserved © Alcatel-Lucent 2009

1 Global optimisation process Global view Mandatory tasks for a new 3G operator


Generic Roll Out Process : RF Design & OPTIMIZATION services Network targets

hypothesis &

Network Radio Planning

assumptions

1a - RF Design guidelines definition

Network Implementation

Network Pre launch Optimisation

2a - Cell Planning

9a - RF Design analysis

3a - Site survey

10a - RF Design preoptimisation

4a - CW Measurement 5a – Propagation Model Calibration 6aGeo.Database Validation

Network Post launch Optimisation

11a - Site Verification

18a - Perf Management

12a - Cluster RF optimisation

19a - Access Benchmark

13a - Network Optimization

20a - Monitoring

14a - Access Parameter Optimization 15a - E2E Optimization

21a - Access parameter customization 22a - E2E audit / Optimization

16a - Application Optimization

7a - Co-siting studies 8a - Spectrum Clearance

17a Configuration analysis

Main Roll Out Step Associated RF Design & Optimization services All these services can be delivered through turnkeys, FTE, workshops or OJTs. Partly covered in design Partly covered in services WS (day2) NDO2009 (day3) All Rights Reserved © Alcatel-Lucent 2009 Workshop OCI – Services Optimisation – December

4

23a - Application audit/ Optimization 24a Densification 25a Troubleshooting 26a - New feature / Load Introduction Support

Optimization Objective 5

• Optimization aims at improving the network performance and/or capacity To improve the system initial performance

Pre--launch phases Pre

To reach acceptance criteria To enhance the subscriber satisfaction

Post--launch phases Post

To increase the number of subscribers

• Methods differ for : – Pre-launch optimization (performed for acceptance, trials) – Post-launch monitoring and optimization (performed for in-service networks)



Optimization Steps 6

Pre-launch optimization

Post-launch optimization KPI Targets Improvement Network E2E optimization LAUNCH

Network Acceptance Network E2E optimization Cluster E2E Optimization Site verification RF Optim

Cluster E2E Optimization

Initial KPI Targets

Tools ALU Initial QoS Benchmark

Improvement is easier before launch


Final QoS Benchmark


Customer

NPO

Call Trace CTb / Ctg

RFO

Mobiles XCAL

XCAP

Scanner

Protocol Analyzer

RF Planning Tool

2 Pre optimisation (design phase)


WCDMA RF Characteristics compared to GSM 8

Coverage quality sensitive to traffic load. Capacity usually limited by Down Link Interference. All neighbor cells are considered as Co Channel interferers => Interference Control Critical for all CDMA-based Technologies. Faster Handoff Processing & Power Control are essential Coverage control should be exercised at the early design stage Optimisation need to be done per cluster area (per group of cell): it means that optimisation must drive site integration

Pilot Ec/Io is the Key metric, not signal Level Coverage & Interference Control !!



Pre-Optimization principle 9

• Pre-requisite. Design threshold for the most constraining service (UL link budget) Network downlink load assumption QoC to be reached (90%, 95%) Pilot power setting and minimum Ec/Io (DL link budget)

• 1st step – initial setting : – Default tilt to be set by environment (Urban 6°,Suburban 4°) – Exceptional tilt to be set according to site height (Tilt 8°if H>=40m)

• 2nd step – simulation tuning: – Maximize the coverage (based on RSCP level) – Minimize the pilot pollution (based on Ec/Io map & overlapping) – Optimize the overlap (overlapping)

• last step : - Global improvement should be check with all tilts implemented Workshop OCI – Services Optimisation – December 2009


Pre-Optimization criteria 10

• Global definition RSCP level >= design threshold (98-100% of the area) Ec/Io>= Ec/Io minimum (90-95% of the area according to QoC) Less than 4 servers in X dB interval from the best (X=4-8 dB) 3rd server level <= design threshold + Y dB (Y= 6-10 dB), to avoid overcoverage (impact in HSDPA optimization) • radio criteria example:

RSCP > - 88 dBm (Indoor Daylight, QoC 90%) Ec/Io > -9 dB (unloaded network) Less than 4 pilots within a 6dB interval from best server 3rd server coverage less than -78dBm



3 Pre-launch optimisation process General view (process/KPI/ressource) Configuration Audit Site Audit Cluster Optimization Network Optimization


3G Optimisation Process Main Steps for pre-launch phase

12

PREPARATION

R99

RF Design Assessment Cluster Definition Acceptance defined PROCESSES Ready RF OK

Cluster Ready

I&C NodeB

Site Verification

Cluster OPTIM.

Providing the best RF & system performance (around Site and inside Cluster/Area) -RF Control -Mobility / Call Performance -Parameter Settings Workshop OCI – Services Optimisation – December 2009

KPI 3G AMR Call drop rate DL FTP Success rate UL FTP Success rate 3G Video Call Success rate


KPIs OK

QoS VERIF.

QoS Audit

HSDPA KPI Throughput FTP DL TV Live Success rate

Pre-Launch Optimization Services : Overall process 13

Step 1: RNP Assessment & Site verification • • •

Access all site configuration and assess RNP settings (tilts, neighbour relations) Project schedule and drive routes defined at this stage for the town. Perform RF field measurements for each new site: Coverage check, crossed feeders Freq/SCs, interference/pilot pollution

Step 2: Cluster RF Optimization • Drive tests done in mobility, on voice services (KPI L1) coverage holes, interference analysis, neighbours' tuning radio link quality of the bearer

Step 3: Town/region EtE Optimization • Drive tests done in mobility for Voice and Static for Data services (KPI L2) Homogeneous QoS End User services tested radio link quality of bearers

Secure Region/Network Launch date with Optimal End User QoS from Day1 Workshop OCI – Services Optimisation – December 2009


Main 3G KPIs pre-launch example 14

Pre-launch RF Optimisation KPIs Test area PROTOCOL TEST KPIs Associated Radio Coverage 95% of Ec/Io samples RF Optimisation report with Drive test (KPIs always part of Cluster coverage maps including Radio quality 95% of RSCP samples Optimisation measurements Ec/Io & RSCP Report) Optimizatio Quality of Services - Drive-test-based KPIs n Level 1 Test area PROTOCOL TEST KPIs Associated Retainability

CS long call voice

Drop call rate

Retainability

CS short call voice

Drop call rate

Accessibility

CS short call

call setup success rate

Mobility

HHO 3G-2G

Service PS UL

PS 128 UL

Service PS DL

PS 384 DL

Service HSDPA Service HSUPA

HSDPA DL HSUPA UL

Optimizatio n Level 2


KPI Value

To be agreed during local RFQ negociation

KPI Value

success rate CS HHO success rate PS HHO


Call setup success rate Average Throughput Call setup success rate Average Throughput Throughput Mobile Cat 12 Mono User-Cell center CQI > 20 Mono User-Cell edge CQI > 15 Throughput Mobile Cat 3 Mono user -Cell center



Optimization Steps 15

Configuration Audit • RF design audit • Spectrum Clearance • System parameters check • Recommended values setting Site Verification • Network integrity verification • H/W troubleshooting Cluster Optimization • Coverage improvement • Neighbouring optimization • Parameters optimization • Cluster fine tuning Network Optimization • Optimization spreading to all clusters • Network fine tuning Workshop OCI – Services Optimisation – December 2009


Configuration Audit 16

Objective •

The objective of the configuration audit is to ensure proper system configuration and functioning in the integrated system.

Checking • • • •

Cell Database: coordinates, # of antennas, antenna type, antenna tilt… RF Design: site and cluster Coverage, SC planning, neighbour list, site configuration Datafill: parameters coherence with Engineering rules and recommendations Spectrum Clearance (Noise floor measurements)

Method • • • •

Cell database information collection and check Coverage prediction maps Datafill check with WiPS Drive tests with spectrum analyzer



Site Verification 17

Objective • The objective of the cell verification is to ensure proper cell site operation in the

integrated system. • The verification is the last check after integration and prior to optimization to detect any residual issues such as cross feeder. • Some voice & PS calls tests are already done during I&C procedure. • It is performed on site by site basis and any issues shall be corrected.

Checking • Softer and soft handoff OK • Scanner received levels are within expected ranges • The UE Tx power is within an acceptable range • Correct scrambling codes and neighbour lists are data-filled • Detection of major issues on antenna tilts and azimuth • Missing SC : compare Top N SC received on the scanner with Active/ Neighbour set

list from the mobile



Site verification 18

Method • • • • • •

Drive tests at approximately 30 percent of the expected cell coverage area Tools : Scanner and Trace Mobile with XCAL Post processing with XCAP and analysis recommending the necessary HW and SW modification due to errors in installation. CS Calls : Call originated on one sector, and move towards other sectors PS Calls : 5 Mbytes (DL) and 1 Mbytes (UL) files are downloaded while driving Other verifications : - cross feeders (while driving) - RSCP (detect any abnormal low value) - Missing SC : compare Top N SC received on the scanner with Active/ Neighbour set list from the mobile

Sector γ Carrier Code γ Power γ


Sector α Carrier Code α Power α Sector β Carrier Code β Power β


Cluster Optimization 19

Objective •

Perform RF design verification,

•

Identify and categorize network or coverage problems

•

Optimize for best performance according to the acceptance criteria and agreed KPI

•

Two phases to be followed

RF Optimization & Parameter Optimization

• RF Optimization – Objective : Reach the optimal configuration of each cell site (height, tilt, EIRP, orientation, antenna type, etc.) Maximize RF QoS (coverage improvement, interference reduction, neighbouring list optimisation) and capacity of the network. – Method : Use of Call trace data and drive test (scanner and UE).

• Parameter Optimization – Objective : Starting from recommended values, the objective is to tune any specific parameters to solve issue related with specific scenarios within the cluster. Set system parameters in order to reach optimal network performance. – Method : Use of KPI dashboard, Call trace and drive test data (scanner and UE).

The cluster optimization process will be detailed in the next section of this document as the E2E optimization will be performed on 2 clusters before being spread on the whole network. Workshop OCI – Services Optimisation – December 2009


Cluster Optimization 20 Define drive test routes

Perform DT intra cluster Check, Analyze, post-process DT raw data Propose HW and parameters change recommendation

Change Parameter is necessary

Aerial Tuning is necessary

Open ticket with when an issue is detected

Ticket Treatment

Change Parameter Process

Operator Agreement

Perform DT until Cluster optimization is completed and target KPI are met

Aerial Tuning

Perform New DT Check, Analyze, post-process DT raw data

Provide KPI progress report on a weekly basis Perform DT until Cluster optimization is complete and KPI targets are met




Main metrics to be used along the process RF Quality indicators Received energy per chip divided by the power density in the

CPICH Ec/Io (dB)

band i.e. it is identical to the RSCP measured on the CPICH divided by the RSSI. Active Set Size

Number of Radio Links established simultaneously by the UE

Application layer Metrics Dropped Call Rate

Number of Calls abnormally terminated;.

Call Setup Success

Percentage of calls established related to the number of

attempted calls Successful Calls

Percentage of calls that could be normally established,

maintained and terminated normally




Drive Tests Runs 1st Run : RF control Scanner + voice (long call) + XCAL Check radio conditions Post processing will mainly be on radio conditions

2nd Run : RF control fine tuning + neigbor tuning Scanner + voice (long call) + CSD (long call) + XCAL Radio conditions + first KPIs (drop / access failure) Post processing will mainly be on access failure / call drops

3rd Run : RF control fine tuning + KPI Mainly pre-KPIs measurements Scanner + Voice, CSD, PS64/384 + XCAL + Call Trace + Probe (if needed) Post processing will mainly be on access failure / call drops / KPI achievement

22




1st Run Post-Processing Mainly with XCAP, Planning tool Check radio coverage (RF Quality indicators) RSCP (Best SC and per SC) Ec/Io (Best SC and per SC) Active set size SC footprint => overshooting Pilot Pollution Comparison with Planning tool

Check Application layers metrics (drop / access failure) Identify problematic areas Determine Antenna changes and Neighbor List change to be done




1st Run Post-Processing: Pilot Pollution Analysis

A map with colour coded symbols (based on the number of sectors beating the threshold) is displayed. 2 possible modes : Nb Sectors above Ec/No threshold Nb Sectors below Ec/No delta threshold (related to strongest Ec/No) Indicates of how many sectors could be in the active set.




1st Run Post-Processing: Coverage Analysis

User selects the Scrambling Code (SC) corresponding to the sector to analyse. 2 possible modes : Best Ec/No All Ec/No instances: Display symbols where Ec/No measurements are reported (regardless of the value) for the selected SC.




2nd and 3rd Runs Post-processing •

Mainly done with XCAP, Actix

•

Main issues to check / solve : Application layer metrics : Access failure / Drop Calls to classify Radio issue Poor Coverage : under thresholds RSCP Configuration • Pilot Pollution (Several servers with equivalent level) => Create a dominant server • Overshooting : strong signal far away from site => Change antenna configuration or add neighbour • Missing neighbour : Good RSCP and Ec/No for best cell in the scanner, bad Ec/No for best cell in the active set. This cell is different to the best one from the scanner. => Check and add neighbour cell




2nd and 3rd Runs Post-processing Radio issue Slow Handover The Ec/Io of a serving SC goes down very quickly, and a defined neighboring SC gets strong very quickly deriving on a fast increase of the DL BLER. Not enough time for active set update => change antenna configuration / change legadd / legdrop parameters. Network (deeper investigation needed) SW issue (neigbhour list, …) HW issue (BTS, transmission, TMA, …) Parameter Configuration (3rd run) Mobile Other



Cluster Optimization Example The following slides show how to analyze a typical dropped call with post drive analysis software

Call Event



28


Example UE Dominant PSC = 85 Active Dominant Ec/No = -25 dB Dominant RSCP = -82.2 dBm

Time of the drop

Scanner Top 1 SC = 164 Top 1 RSCP = -52 dB Top 1 Ec/No= -4.9 dBm




Example This slide shows the UE dominant active Ec/No at the time of the drop (value= -25 dB) UE is moving left to right

UE Plot Active Dominant Ec/No




Example This slide shows the UE dominant PSC at the time of the drop (value= 85)

UE Plot Dominant PSC




Example This slide shows the scanner Top 1 SC (164) at the time of the drop

Scanner Plot Top 1 SC




Example This slide shows the scanner Top 1 RSCP (-52 dBm) at the time of the drop

Scanner Plot Top 1 RSCP




Example The UE is on PSC 85 and driving toward the cell with PSC 164. As the UE gets closer to the site with PSC 164, the Ec/No degrades to a point where the call drops. The UE is on PSC 85 at Ec/No = -25 and the scanner is showing the best SC is 164 at -52 dBm (-4.9 Ec/No). When the UE drops, the call is re-established on PSC 164. This is a typical scenario for a missing neighbour. Reviewing the neighbour list for the cell with PSC 85 shows that the cell with PSC 164 is not declared as its neighbour.



Network Optimization 35

Objective • The objective of network optimization is to optimize the sum of all clusters covering an entire region

Checking • Agreed KPI

Method • Verify network performance covering several clusters simultaneously • Perform the final fine-tuning of the network settings paying special attention to the inter-cluster sites. • Collect metrics and statistics: KPI dashboard • Conduct drive tests on selected clusters or key routes in case of KPI degradation.



4 Post-launch optimisation process


Cluster Optimization Process 37

Monitoring (KPI dashboard) Accessibility Retainability Mobility Capacity Quality

Drive Test Scanner trace UE Trace: XCAL IMSI trace: CTb

Yes

Yes Target DT KPI reached

No compulsory optimization

No

No

RF Optimization Coverage improvement Interference reduction Neighbouring optimization Capacity


Target Dashboard KPI reached

HW Change/Addition Cross Feeder iCEM MCPA E1 ...


Parameters Optimization Accessibility Call Admission Control Radio Resource Management Mobility Iub Efficiency HSDPA

Drive Test Based Optimization 38

Define drive test routes

Perform DT intra cluster Check, Analyze, post-process DT raw data Propose HW and parameters change recommendation

Change Parameter is necessary

Aerial Tuning is necessary

Open ticket with when an issue is detected

Ticket Treatment

Change Parameter Process

customer Agreement

Aerial Tuning

Perform New DT Check, Analyze, post-process DT raw data

Provide KPI progress report on a weekly basis Perform DT until Cluster optimization is complete and KPI targets are met



Perform DT until Cluster optimization is completed and target KPI are met

Counter Based Optimization

39

Monitoring Retainability

Accessibility Includes all the steps allowing the user to access a particular service

Defines the ability of the network to provide reliable service to the end user.

Counter based KPI

Mobility

Quality Defines the ability of the network to ensure UE’s calls continuity.

Defines the ability of the network to provide quality transmission of the user requested services

Traffic Defines the traffic volume seen on the network in downlink and in uplink

Monitoring concerns all aspects of QOS Workshop OCI – Services Optimisation – December 2009


Counter Based Optimization

40

Check the daily and weekly KPI dashboards

Yes Target KPI reached

No compulsory optimization

No

Check the Worst Cell List

Perform DT in addition to the check of the KPI dashboard

Global Issue or on specific cells

Troubleshooting: additional metrics/counters, CTg, DT (UE/CTb)

No Root cause found

Open a ticket

Yes Apply corrective actions: HW or parameters change



Ticket treatment

5 Optimisation tools


Drive Test Measurement Chain OUTDOOR

42

GPS

Magnetic Antenna

PC with

INCAR

Acquisition SW

Dual Frequency Scanner R99 UE: HSDPA UE: Option datacard Cat. 12 / Cat. 6 (Qualcom chipset)



Optimization Tools Set

43

IP

OAM Call Trace

RNC Iu Iub

RFO+DT post processing SW

NodeB

Workorders

ALU ALU ?

Protocol Analyzer

WiPS

ALU Data Collection SW

Counters NPO

A9155 Scanner UE ALU Workshop OCI – Services Optimisation – December 2009

ALU


ALU

Call Trace 44

Call Traces are activated on the OAM GUI. The log files are stored on the Performance Server. Trace all calls of one or several UEs

Most used Call Traces

CTa

Activated from Core Network

Trace all calls of one specified UE CTb

Activated from Access Network

Trace a % of calls over the NW on a specific cell or group of cells. CTg

CTn

OTCell

Activated from Access Network Trace a % of calls (only neighbouring optimisation related messages) over the NW on a specific cell or group of cells Activated from Access Network Post-Processing with WQA Tool Trace data related to all elements (events) handled by an object of the network : A cell or group of cells Iu CS interface Iu PS interface One Neighbour RNC or a group of neighbour RNC One RNC Activated from Access Network



45

www.alcatel-lucent.com



46

Backup tool



Call Trace B: CTb 47

A Call Trace B session provides data related to all calls initiated by a specified UE during the session activation. There is one and only one UE traced in a CTb session. Therefore if n mobiles are to be traced simultaneously, then n sessions are created. The UE identity, specified at the OMC-R when creating a CTb session, may be one of all choices of the “Initial UE identity” IE supported by the RRC protocol specified in 3GPP specification 25.331: TMSI,P-TMSI, IMSI, IMEI.



Geographical Call Trace: CTg 48

A Geographic Call Trace (CTg) session allows to trace a % of the calls established within a geographical area in the UTRAN specified during the session activation. The geographical area in the UTRAN is a list of geographical areas in specified RNSs. RNSs are defined by their RNC Id. The geographic area in a RNS may be a cell, a set of cells from the RNS, all the cells in the RNS. There are therefore no UEs identified in a CTg session creation. Calls may be traced within a CTg as soon as the trace session is activated and exists.



RFO (RF Optimizer) 49

Post-processing Tool to analyse data log files from UE acquisition tools and Call Trace. It supports Nemo traces and is able to synchronize them with CTb traces.



NPO (Network Performance Optimizer) 50

ALU’s Performance Management Solution. Allows to monitor the network performance, detect network issues and troubleshoot them through the counters. Collects and stores for long periods of time all the UTRAN counters. Includes metrics (indicator based on several counters) defined by Alcatel-Lucent. The counters are available at RNC level and/or at FDDCell level. Easy report generation (drag & drop). Pre-defined reports available. Intersystem Cell Reselection 1200000

100.00% 98.00%

1000000

96.00% 94.00%

800000

92.00% 600000

90.00% 88.00%

400000

86.00% 84.00%

200000

82.00% 0 15-mar-06

80.00% 22-mar-06 Attempts

29-mar-06 Success

Success Rate



ALU RF Optim Process

Recommend Documents