You are on page 1of 32

THE PRINCIPLES OF DRIVE TEST REALIZATION AND RADIO-PARAMETERS

OPTIMIZATION IN THE PROJECT OF MOBILE RADIO BASE STATION NETWORK MODERNIZATION


Predrag Radic Vladica Kovacevic Ivan Milovic
Kodar Engineering Ltd. Belgrade, November 2011.

PURPOSE AND GOAL OF DRIVE TESTS AND OPTIMIZATION


IN THE PROCESS OF NETWORK MODERNIZATION
RAN (Radio Access Network) Swap:
New BSC/RNC controller installation New OSS system installation Base stations swap

+ Possibly replacement:
Transmission system PS Core network (SGSN / GGSN) CS Core network (HLR, MSC, MGW)

PURPOSE AND GOAL OF DRIVE TESTS AND OPTIMIZATION


IN THE PROCESS OF NETWORK MODERNIZATION
KPI statistics of the existing network on the cluster level (Key Performance Indicatiors)

RAN (Radio Access Network) Swap:


New BSC/RNC controllers installation New OSS system installation Base stations swap

KPI statistics of the new installed network on the cluster level

User experience KPIs of the existing network Link availability, speech quality, Web services (download, upload, real-time streaming, ...)

+ Possibly replacement:
Transmission system PS Core network (SGSN / GGSN) CS Core network (HLR, MSC, MGW)

User experience KPIs of the new installed network

PURPOSE AND GOAL OF DRIVE TESTS AND OPTIMIZATION


IN THE PROCESS OF NETWORK MODERNIZATION
KPI statistics of the existing network on the cluster level (Key Performance Indicatiors) KPI statistics of the new installed network on the cluster level

RAN (Radio Access Network) Swap:


New BSC/RNC controllers installation New OSS system installation Base stations swap

User experience KPIs of the existing network Link availability, speech quality, Web services (download, upload, real-time streaming, ...)

RNO & DT

+ Possibly replacement:
Transmission system PS Core network (SGSN / GGSN) CS Core network (HLR, MSC, MGW)

RNO & DT

User experience KPIs of the new installed network

Achieve fully network functionality during swap

Notice and repair every hardware/ software mistake occured during swap

Provide the same or better KPI statistics

Provide the same or better user experience indicators

PURPOSE AND GOAL OF DRIVE TESTS AND OPTIMIZATION


IN THE PROCESS OF NETWORK MODERNIZATION
NEW RAN NETWORK:

EXISTING RAN NETWORK:


Mainly at least a 10 years of exploatation, with upgrading of Base stations and controllers from different vendors, changing stuff in RNP/RNO team etc.

Perfectly clear network insight with:


1. RAN Swap 2. Onsite hardware state documentation Consistency and documentations of Data and Radio-parameters Documentations and relevancy of KPI Data Detailed network ilustration obtained by Drive tests

3. 4.

Among network hardware/ software modernization itself, swap is excellent opportunity to fully adjust and advance network, documentation, databases, etc.

PROCESS OF SWAP REALIZATION


General Preparation Cluster SWAP Preparation
Cluster SWAP

Cluster Optimization

Cluster Verification

1. 2. 3. 4. 5.

General preparation Cluster swap preparation Cluster swap Cluster optimizaton Cluster verification

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

GENERAL SWAP PREPARATION


Radio parameters description Network parameter dump Cell Traffic Data Performance counters description

KPI description (formulas)


DT Cell data

Transmission Network Design

Radio parameters description

Radio parameters mapping

Performance counters description

Defining of target KPIs as well as formulas for calculating

Cluster defining

Period of intense communication and information exchange with network operater


Establishment of crucial conditions necessary for successful swap

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

GENERAL SWAP PREPARATION


Radio parameters description Network parameter dump

VARIOUS HANDOVER ALGORITHMS: Handover parameters: hysteresis, offsets, thresholds etc VARIOUS HCS ALGORITHMS: Cell level and priority parameters RADIO-UNITS WITH VARIOUS CHARACTERISTICS AND POWER SETTINGS: In order to achieve equal TOC power ALL OF THE OTHER VENDOR-SPECIFIC PARAMETRS AND ALGORITHMS.......

Radio parameters description

Radio parameters mapping

Global Default Cell Non-default Relation Direct mapping Conversion

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

GENERAL SWAP PREPARATION


Radio parameters description Network parameter dump Cell Traffic Data Performance counters description

KPI description (formulas)


DT Cell data

Transmission Network Design

Radio parameters description

Radio parameter mapping

Performance counters description

Defining of target KPIs as well as formulas for calculating

Cluster defining

OSS KPIs defining:


KPIs proper choice KPI formulas mapping between two networks Time frame choice (typically 1 3 weeks) and time samples during the day for KPIs collection (busy hour, 24h, etc) Type of counting KPIs: summary or averaged per day

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

GENERAL SWAP PREPARATION


Radio parameters description Network parameter dump Cell Traffic Data Performance counters description

KPI description (formulas)


DT Cell data

Transmission Network Design

Radio parameters description

Radio parameters mapping

Performance counters description

Defining of target KPIs as well as formulas for calculating

Cluster defining

Defining of target DT KPIs

Defining of DT scenario

Defining of DT report content

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

CLUSTER DEFINING

Cluster collection of neighboring sites; usually 15 25 All DT and RNO activities during the swap are done on the cluster level During the cluster definition process the things that need to be taken into acount:

Terrain (physical borders) Traffic distribution (should be similar on all of the clusters) Transmission lines (all of the sites should be on the same transmission line) Logical urban zones borders Radio controller (BSC/RNC) responsibility

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Stability

DEFINING OF TARGET DT KPIS AND SCENARIOS


List of DT KPIs and services Static test participation determining Drive test and Static test scenarios

In order to achieve high-quality Drive test it is important to define appropriate DT scenario and KPIs to present in report
Mutually test impact:
If done at the same time some tests can impact each other, for example: HSDPA DL test has negative impact on the scanned Ec/Io!

Dual mode test


Consider dual mode test because: Defining UMTS service zone Defining samples where inter-RAT handovers occur It is necessary to quit the call when inter-RAT handover happens and then after cell resellection to UMTS to establish a new call

Static points
Traffic tests can be done in motion or in static points. Good practice is to do motion traffic tests at the beginning of the project, and later to switch to static tests in order to utilise resources .

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

DT SCENARIO EXAMPLE
Description of the test used to collect KPIs LIST of exported KPI / IE Test software sequence description

Test type Test Description

KPI
Strongest Scanned GSM 900 BCCH ARFCN Strongest Scanned GSM 1800 BCCH ARFCN Strongest Scanned GSM 900 RxLev Strongest Scanned GSM 1800 RxLev RxLev Sub / Full BCCH ARFCN RxQual Sub / Full SQI Time Advance Call Setup Success Rate Call Drop Rate Handover Success Rate FTP UL Session Application Throughput FTP UL RLC Throughput FTP DL Session Application throughput FTP DL RLC Throughput PDP Context Activation Success Rate

Device number
MS1

Device type
Scanner

Sequence description
1.) GSM Scan All Operator's BCCH / BSIC

Drive test

GSM RF Coverage

Drive test

GSM Voice Short Call

MS2

Phone

1.) UE locked on 2G 2.) 50s duration + 10s idle

(E)GPRS FTP Upload (E)GPRS FTP Download Drive test

(E)GPRS PDP Context Average PDP Context Activation Time PS Session Drop rate Ping Delay (E)GPRS Ping Service Ping Number Ping Packet Loss

MS3 / DC3

Modem

1.) UE locked on 2G 2.) 60 s FTP Upload 3.) 10s wait 4.) 60s FTP Download 5.) 10s wait 6.) Ping 10 times 7.) 10s wait 8.) Repeat

13

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

DT EQUIPMENT CONFIGURATION
On the basis of determinated scenario, the final equipment configuration will be defined Hardware problems impact on results have to be minimized:
The differencies in transmitting-recieving characteristics of different devices Possible degradation of the receiving characteristics of device during the time!

The same test set have to be used(reference, validation, acceptance) on the cluster
On the test set device allocation have to be the same

Periodically test of the received signal strength for all of the the devices

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

DT REPORT CONTENT DEFINING INFORMATION ELEMENTS


DISPLAY

DT report form is designed especially for the project in accordance with agreed IE and KPIs with operater.
Standard DT results display:

RxQual - 2G
Interval 0 <= x < 5 5 <= x < 6 6 <= x < 7 7 <= x < 7 PDF [#] 24564 141 57 33 PDF [%] 99.1 0.6 0.2 0.1 CDF [#] 24564 24705 24762 24795 CDF [%] 99.1 99.6 99.9 100.0

MAP

Samples (#)

24,795

STATISTICS

Average

0.2

GRAPH

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

DT REPORT CONTENT DEFINING INFORMATION ELEMENTS


DISPLAY

KPI values are obtained:


1.

Statisticaly, from information elements, where value may be: a) Averaged from all of the samples b) Sample percentage with the IE value above/under defined border value
Reference Average Rate value value -74.5 98.60 % 0.2 99.60 % -81.3 89.90 % -5.4 99.40 % Acceptance Average Rate value value -75.3 98.30 % 0.2 99.50 % -80.9 90.20 % -5.1 99.60 %

Mode GSM GSM UMTS UMTS

KPI (unit) RxLev Sub (dBm) RxQual Sub CPICH RSCP (dBm) CPICH Ec/Io (dB)

Formula (#samples_RxLev>=-85dBm)/(#samples) (#samples_RxQual<=5)/(#samples) (#samples_RSCP>=-95dBm)/(#samples) (#samples_Ec/Io>=-11dB)/(#samples)

Pass?

2.

Mathematical expression of events:


Reference Acceptance Failures Rate Failures Rate count value count value 12 97.30 % 8 97.90 % 4 98.80 % 3 98.90 % 2 99.00 % 2 99.10 % 0 100.00 % 0 100.00 % 0 0 4 2 0 100.00 % 100.00 % 98.70 % 99.30 % 100.00 % 0 0 0 1 0 100.00 % 100.00 % 100.00 % 99.50 % 100.00 %

Mode GSM GSM GSM GPRS GPRS GPRS UMTS UMTS UMTS

KPI Call setup success rate Call drop rate Handover success rate PS Attach success rate

Formula 100 * (#Call Setup / #Call Attempt) 100 * (#Dropped Call / #Call Setup) 100 * (1 - Handover failures / Handover attempts ) 100 * (1 - #PS Attach Failure / #PS Attach) (1 - #PDP Context Activation Failure / #PDP Context PDP activation success rate Activation) PDP session drop rate 100 * (1- PDP dropped connections / #Session Start) Call setup success rate 100 * (#Call Setup / #Call Attempt) Call drop rate 100 * (#Dropped Call / #Call Setup) 100 * (1 - #Handover From UTRAN failure / Handover fro UTRAN success rate #Handover From UTRAN)

Pass?

16

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

DT REPORT CONTENT DEFINING ADVANCED


1. Unwanted events detailed display
# Lat Lon Time Logfile Event Event Info Block type: No Immediate Assignment. , call time: 10297 ms Last Best Last Best Server Name Server RSCP Last Best Server EcIo Last UE Dist from TxPower Serving Cell

1 44.49473 20.25293 15:43:06 Test_01

Blocked Call

Cell_1

-84.3

-11.3

33

2. Geographical display:

3. Layer 3 signalization display (immediatelly before bad event)


UE RRC Connection Request RRC Connection Setup Complete RRC Connection Setup RNC

CM Service Request

Authentication Request

17

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

DT REPORT CONTENT DEFINING ADVANCED

It is possible to present additional analysis of the unwanted network performances, like: Pilot Pollution Missing Neighbour

The data are combined from scanner (neighbour signal strength and quality) and from test phone (BA list) for GSM For detailed analyse it can be used predefined Cell & Neighbors Data

Ping-pong handovers
Polluted samples (#) 1814 1675 1122 1003 773 708 696 650 638 628 520 513 493 482 Detected Samples (#) 15986 8557 7482 7136 6094 5450 4718 4601 3660 5899 3071 3550 3714 5138 Polluted samples (%) 11.3 19.6 15.0 14.1 12.7 13.0 14.8 14.1 17.4 10.6 16.9 14.5 13.3 9.4

Polluter Cell Name HN3395 HN3394 TI3345 TI3346 TI3340 TI3147 TI3344 TI3050 HN3398 HN3430 HN3422 HN3396 HN3390 TI3148

SC 299 291 227 235 243 178 219 18 315 435 419 307 267 186

Highest Polluter Cells


Polluted samples (%)
80.0 60.0

40.0
20.0 0.0

19.6
11.3

15.0 14.1 12.7 13.0 14.8 14.1 17.4

10.6

16.9 14.5 13.3

9.4

HN3430

HN3422

HN3396

HN3390

HN3395

HN3394

HN3398

TI3148

TI3345

TI3346

TI3340

TI3147

TI3344

TI3050

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

PREPARE FOR SWAP AND CLUSTER SWAP


Cluster freeze period start Reference OSS KPI Collection Cell Name Modification Table Cell & Neighbors Design Data

DT Cell data

Cell Design Data

Cluster swap start

New cells configuration scripts generation

Swapped Sites List


BSC/RNC Dumps

Non-default cell parameter data

OUTGOING neighbouring relations scripts generation

INCOMING neighbouring relations scripts generation

Definisanje DT rute

Reference drive test

Single site verification test

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

DT ROUTE DEFINING

DT route should cover all of the cells serving zones as well as main roads Usual route length: 5 15 km per site about 200 km per cluster If cell level analyse is needed, for example Missing Neighbors and Pilot Pollution in urban zones, route should cover the majority of streets because of statistically relevance of result data.
DT route have to be identical in every repeated test. Because of handover hysteresis effect there is a need for identical route direction in every of repeated tests.

Exceeding of the service zone (UMTS test): Route have to be inside expected UMTS service zone, in order KPI statistics to be relevant

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

REFERENCE DRIVE TEST


It is performed after cluster parameters freezing announcement in order to collect reference DT KPIs

Collection of DT KPIs on the existing network

Reference DT

Notice of mistakes in the DT Cell Data file

Notice of hardware mistakes/problems which can be solved during the swap-a (mixed feeders, sectors)

Reference DT Report

Notice of problematic areas where special attention should be payed after the swap

Exact day (working day/weekend) and period during the day when DT is performed have to be repeated on other test on that cluster, in order to proper compare the results

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

NEIGHBOURING RELATIONS COMMAND


SCRIPTS GENERATION

Actions during the swap of base station:


Reconfiguration of neighbouring relations from and to swapped cell Define/change definitions of external cells

Important to provide:
Independency from previous system; no need for any changes Fully functionality just after new cell become active Independency from swap cell order and dynamics

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

RELATIONS CONFIGURATION SCRIPT MAKER


Software tool internally developed for automatically generation of command scripts and network parameters consistency checking Developed for needs of Telenor Renovation project, adaptable for any similar project Provides command generation and control if Cell & Neighbors neighbouring relations parameters are correctly set
Cell Name Modification Table

Design Data

Cluster sites swap

Swapped Sites List BSC/RNC Dumps

OUTGOING neighbouring relations scripts generation

INCOMING neighbouring relations scripts generation

Previous detailed validation of input data External cell adding/ modification Defining of externall cell index in order with mobile operators demands Cell names harmonization Neighbouring names adding

External cell modification/ removing Redirection of neighbouring relations from old to new cell

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

SCRIPT MAKER CHECKING


Initial script generation

Swap script generation

Check if external cells parameters are correlated with initial parameters Check if neighbouring relations parameters are correlated with initial parameters Check if parameters with no default values are correct set
Checkings

Unexisting cell finding


Asymmetrical neighbouring relations finding ...

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

OPTIMIZATION CLUSTER PROCESS


Daily OSS KPI monitoring

Optimization and radio parameters tuning

Freezing cluster parameters

Functionality check and urgent problem solving


Change Request

Determination if reference KPI values are reached

Acceptance OSS KPIs collection

Acceptance OSS & DT Reports

Validation drive test

Troubleshooting drive tests

Acceptance drive test

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

OPTIMIZATION CLUSTER PROCESS


Daily OSS KPI monitoring

Basic functionality check


Functionality check and urgent problem solving

Parameters check
Basic cell parameters Neighbouring relations

Hardware check
Critical alarms High VSWR Crossed/mixed feeders/cells TMA device check

Voice traffic Data traffic Handovers

Validation drive test

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

CROSSED CELLS / FEEDERS


1. MIXED CELLS (TX BRANCHES) 2. TWO SC (BOTH TX BRANCHES) IN
THE SAME AZIMUTH

3. MIXED RX BRANCHES

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

OPTIMIZATION CLUSTER PROCESS


Daily OSS KPI monitoring

Optimization and radio parameters tuning

DT result analysis
Cells with unstable service zone Zones with quality degradation Zones with reduced data traffic

Change Request

Overshooting cells
Bad events analyzing Layer 3 signaling detailed analyze

Validation drive test

Troubleshooting drive tests

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

OPTIMIZATION CLUSTER PROCESS


Daily OSS KPI monitoring

Optimization and radio parameters tuning

Optimization and radio parameters tuning


Congestion problems solving and traffic optimization: Cell traffic distribution Technology traffic distribution Interference problems solving Neighbour optimization The worst cell analyze and problem solving

Change Request

Validation drive test

Troubleshooting drive tests

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

TRAFFIC DISTRIBUTION OPTIMIZATION

Basic tasks:

Solving the new noticed traffic congestion Provide optimal traffic balance

Comparing traffic distribution per cell before and after swap visual analysis and possible problem noticing Detailed analyse of KPIs, serving zones comparision, ...

Possible actions (GSM):


Cell Reselection parameters change Handover parameters change Direct Retry to switch on Half Rate Treshold tuning Hardware TRX reallocation between cells

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

NEIGHBOUR OPTIMIZATION

OSS tools enable finding of missed 2G neighbours from BA2 list, as well as all of missing 3G neighbours Drive test analyse: Enables finding of missed ARFCNs in BA2 list, as well as all of missing 3G neighbours

Map display of missing neighbours, geographical presentation is desirable

Automated scripts generation by using of Script Maker

General Preparation

Cluster SWAP Preparation

Cluster SWAP

Cluster Optimization

Cluster Verification

CLUSTER VERIFICATION

Cluster optimization
(couple weeks)

Preliminary Acceptance Certificate

Cluster verification
(couple months)

Final Acceptance Certificate

SWAP SUCCESFULLY FINISHED

You might also like