ACCESSIBILITY

RETAINABILITY
3G
SOS

MOBILITY

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 2 (46)

INTEGRITY
› INTRODUCTION

› Mobility Kpis & formulae

› TROUBLESHOOTING
– Possible Causes

– L1 Checklist

– L2 & L3 TROUBLESHOOTING

› Case Studies

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 3 (46)

 Introduction

Mobility is defined as the ability of a user to move between neighboring

UMTS and GSM Cells and retain its requested service (CS or PS) once connected.
Two major components of mobility KPI are Soft & IRAT Handover Success
Rate

Soft Handover
Soft Handover is controlled through the events 1a, 1b, 1c & 1d:
 1a: A new candidate for the active set enters reporting range
 1b: A cell in the active set leaves the reporting range
 1c: A cell not in active set becomes stronger than a cell in the active set
 1d: Any cell becomes better than the best cell in the active set

Majority of the SOFT/SOFTER Handover failures include:

 Failures due to interference, poor areas of dominance
 Primary Scrambling Code collision
 Missing neighbors
 Neighboring sites Hardware issues
 software bug (cell locked, sub channels unlocked)

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 4 (46)

 I RAT Handover

IRAT handover process includes the following two sub-processes:

• Relocation preparation
• Handover implementation

Relocation preparation Failures are mainly due to following reasons:

• The 2G resources are not enough.
• The CN parameters are not configured reasonably.
• The configurations of GSM neighboring cells are not consistent with actual
parameters.

Handover implementation failures are mainly due to following reasons:

• Incorrect parameter configuration of neighboring 2G cells
• There exists side-channel interference in 2G cells
• Improper setting of neighboring threshold

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 5 (46)

 Formulas
SOFT HANDOVER SUCCESS RATE (%)
=100*(pmNoTimesRlAddToActSet /
(pmNoTimesRlAddToActSet + pmNoTimesCellFailAddToActSet))

CS-IRAT HANDOVER SUCCESS RATE (%)

=100*(pmNoSuccessOutIratHoSpeech + pmNoSuccessOutSbHoSpeech +
pmSuccLbhoSpeech) / (pmNoAttOutIratHoSpeech + pmNoAttOutSbHoSpeech +
pmAttLbhoSpeech))

Note: In W14A, LBHO & SBHO Counters (pmSuccLbhoSpeech, pmAttLbhoSpeech &

pmNoSuccessOutSbHoSpeech , pmNoAttOutSbHoSpeech) added in CS IRAT
Calculation. While pmNoAttOutIratHoMulti & pmNoSuccessOutIratHoMulti are
removed.

PS-IRAT HANDOVER SUCCESS RATE (%)

=100*(pmNoOutIratCcSuccess / pmNoOutIratCcAtt)

INTER-RNC HANDOVER SUCCESS RATE (%)

=100*(pmSoftSofterHoSuccessIur / pmSoftSofterHoAttemptIur)

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 6 (46)

 troubleshooting
• 3G-3G Handover
• IRAT CS & PS

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 7 (46)

 Troubleshooting

Mobility

3G-3G HO IRAT

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 8 (46)

 Troubleshooting
Possible Reasons for 3G-3G HO Failures

• Interference
• Pilot Pollution
• Overshooting of nearby cells
• Low coverage area
• External Interference

• PSC Collision
• Two or more nearby Cells having same PSC

• Missing Neighbour Definition

• Improper cell relations resulting into failed HO

• Neighbouring Site E1/Hardware Issue

• Attempts to nearby site failing due to E1/Hardware issue

• Improper Definitions
• Improper External Definitions in RNC/MSS

• Congestion on neighboring site

• Limited resource neighboring site will lead to failures

• On Utran Cell Locked, Channels Unlocked

• Attempts to the cell will take place based on unlocked cell, but will not succeed as cell is
locked
E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 9 (46)
 Sanity Check Checklist

NodeB Current Alarms

20/40 W
Site Configuration maxDlPower
RBS SW Version
NodeB alarm logs
dlCodeAdm
hsdpauseradm
MaxTransmissionPower
primaryCpichPower
Power offset
Utrancell Pwradm
qQualMin
qRxlevMin
qualMeasQuantity
sHcsRat
sRatSearch

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 10 (46)

 troubleshooting
• Soft Handover

• IRAT CS & PS

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 11 (46)

 Troubleshooting

Mobility

3G-3G HO IRAT

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 12 (46)

 IRAT
Sanity Check Checklist

• External Cell Definition – Cell defined in RNC/BSC

• Outer Definition – Cell defined in MSS/MSC
• SGSN Definitions – LAC/RAC Def
Definition • RNC/MSS Definitions – Correct RNC/MSS(for new
Check nodes)

• Blocking
• Hardware Failure
• High ICM band
Hardware • E1 Issues
Check

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 13 (46)

 Possible Causes & Counters

CS IRAT HO Failure

Preparation
Execution Failure
Failure

Physical Channel not due to Physical Failure due to UE

Preparation failures - Failure - Channel Failure - rejection -
pmNoFailOutIratHoSp pmNoFailOutIrat pmNoFailOutIratHo pmNoFailOutIrat
eechGsmFail HoSpeechReturn SpeechReturnOldCh HoSpeechUeReje
OldChPhyChFai NotPhyChFail ction

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 14 (46)

 Optimization and
improvement

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 15 (46)

 Neighbor list tuning
Troubleshooting

Optimum IRAT neighbor list is crucial for IRAT performance due to SIB 11 size limit.

Addition : Need to add missing IRAT neighbors based on 2G-2G ho stats, distance, orientation.
Collocated 2G cell should always be defined as IRAT neighbor

Deletion :Unnecessary IRAT relations having very less attempts or geographically far off having
less success rate need to be removed to avoid missing important IRAT neighbor in MN list.

Frequency : Addition/deletion activity needs to be performed at regular interval to accommodate

new rolled our sites in IRAT relation definitions.

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 16 (46)

 New Features
L3 Troubleshooting

Following Set of Features can be considered for improving Mobility

Number Title Basic / Optional From To

FAJ 121 406 Directed Retry to GSM Optional Off On

FAJ 121 435 Inter Frequency Load Sharing Optional Off On

FAJ 121 799 Service based Handover to GSM Optional Off On

FAJ 121 1356 Load Based Handover to GSM Optional Off On

FAJ 121 1402 Load Based Interfrequency Handover Optional Off On

FAJ 121 1468 Non-HSPA Interfrequency loadsharing Optional Off On

FAJ 121 2135 Soft Handover parameters on Cell level Optional Off On

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 17 (46)

 Case Studies

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 18 (46)

 Optimization by L1 & L2

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 19 (46)

 GSMTHRESH3A & TCELLCHANGE

2 3 4

Default Old New

Parameter Type
Value Value Value

2 GSMTHRESH3A RNC -95dBm -95dBm -90dBm

1
TCELLCHANGE RNC 20Sec 20Sec 30Sec

3 4

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 20 (46)

 GSMTHRESH3A & TCELLCHANGE Contd…

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 21 (46)

 GSMAMOUNTPROPREPEAT Contd…

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 22 (46)

 IRAT HOSR Degradation

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 23 (46)

 SHO SUCCESS RATE DEGRADATION

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 24 (46)

 GSL Devices

PS I-RAT from 3G to 2G involves getting

resource in 2G network. In case of limitation,
like high RPP Load or High GSL utilization, PS
I-RAT failure will increase.
It can be seen that number of devices GSL
devices reserved in BSC are less than number
of devices reserved on Abis or Air Interface.
Hence causing limitation in BSC

Improvement observed in PS-IRAT

by increasing number of GSL
devices
E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 25 (46)
 IndividualOffset
It is observed that areas with higher
overlapping have higher Soft Handovers.
Resulting in higher signaling. For reduction
of SHOs IndividualOffset tuning can be
used.
In this case cluster of 6 sites was chosen for
parameter tuning

Tuning of this IndividualOffset shows that

value of -1.5 is showing improvement of
12.6 in Soft Handovers

Also there was reduction of Soft

Handovers at neighboring sites
due to reduction of overlapping

There is no major impact on Call Drops & CSSR

For first implementation of SHO reduction, antenna downtilt should be the first choice
E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 26 (46) of optimisation.
 Incorrect Definition(s)- 4 case studies
New launched site was defined twice in the RNC. One as internal cell
and one as an external cell due to which the handovers were failing.

On deleting the
External Def
extra external
definitions from
the RNC. The
Internal Def
handover started
happening

Another similar case where Iur link definition was wrong

The Site belongs to the 1602 RNC but the definition was made with
the 1601 RNC.
On correcting the
definitions. The
handover started
happening

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 27 (46)

 Definitions Cont…
Similarly some other incorrect definitions will also make IRAT attempt to fail.
Some of these are shown below:

Incorrect RA definitions in SGSN:

Routing Area is defined in SGSN, in the above case this definition was
incorrect.
E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 28 (46)
 Definitions Cont…
Incorrect ARFCN:
In this case drop rate had
increased suddenly. It can be
observed in drop reasons that
most of the drops are occuring
due to SOHO & IFLS

During definition check incorrect

ARFCN definitions was observed

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 29 (46)

 Cell/Channel Locking

It can be seen that RL Addition Attempts are

happening but Success is zero. But at the
same time co-site relations are okay.
Site WIAP1 was DISABLED but some of the
channels were UNLOCKED(FACH, RACH,
etc.). The neighboring cells were attempting
SOHO to them. Such SOHO attempts lead to
SOHO failures.

Improvement post Locking

complete site
E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 30 (46)
 usedFreqThresh / serviceOffset
• A trigger for IRAT HO will be raised for following conditions:
• CPICH RSCP < usedFreqThresh2dRscp + serviceOffset2dRSCP
• CPICH ECNO < usedFreqThresh2dEcno + serviceOffset2dECNO
• For cells in the core of city, the values can be set lower as IRAT can be avoided as WCDMA
coverage will be good. For city boundary it can be set higher.
• Following values were used in this case:
o usedFreqThresh2dRSCP : on cell basis, ranging from -101 to -107(from -105)
o usedFreqThresh2dEcno : on cell basis, ranging from -10 to -13(from -13)
o FDDQMIN (in GSM Cells) : 6(from 7)

Improvement was observed CS IRAT, PS IRAT, CS DCR

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 31 (46)
 usedFreqThresh / serviceOffset

Improvement was observed CS IRAT, PS IRAT, CS DCR

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 32 (46)
 Q&A
E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 33 (46)
 Preparation Phase
It can be seen that pmNoAttOutIratHo is pegging only after preparation phase is
complete. If there is limitation on 2G core side, this counter will not peg. If the
limitation is faced after preparation phase the attempt will be counted and will be
failed also.
Hence in this case even after being critical limitation of GSL devices, 93% Irat
success was observed. But it can be seen that after increasing the devices the
number of attempts has almost doubled.

Preparation
phase on
2G Side

E/// 3G Mobility SSP | GRAN/WRAN/D | 2014-07-22 | Page 35 (46)