RANOP1_5 - ISHO.pdf

1 © Nokia Siemens Networks Presentation / Author / Date

3G RANOP 1 Module 5 –Inter System Handover (ISHO) Optimisation


Module 6 – Inter System Handover Optimisation

Objectives

• Review the 3G <> 2G Cell re-selection process and

parameters

• Understand the key areas of optimisation for 3G <> 2G

Cell re-selection process

• Review the Handover Process & Compressed Mode

• Understand 3G ISHO Service Optimisation (AMR and PS

Data)


ISHO Optimisation - Agenda

• 3G <> 2G Cell Re-selection Process

• 3G <> 2G Cell Re-selection Analysis

• Handover Process & Compressed Mode

• 3G ISHO Analysis

• 3G ISHO Service Optimisation


Optimisation Process- Goals

• Optimum Cell Re-selection and ISHO performance should satisfy the following

requirements;

• Optimum setting of trigger thresholds to;

• Ensure 3G<>2G transitions are triggered at the correct time to prevent call set

failures and call drops due to 3G coverage/interference

• Avoid unnecessary transitions and minimising associated signalling load

• Satisfy traffic management strategy (i.e maximise ‘Time on 3G’ if required)

• Ensure selection of good 2G GSM target cell in terms of radio conditions (best cell)

• Minimise time delays involved in the 3G<>2G reselection and ISHO processes

• Maximise end user experience

• Achieving optimum performance requires;

• Detailed understanding of the processes and associated parameters

• Field investigation analysis to benchmark current performance

• Trials to investigate potential changes to optimise performance

• Optimum 3G<>2G neighbour plan


Cell Reselection 3G -> 2G Procedure

• Whilst camping in a 3G cell the UE performs intra-frequency, inter-frequency, and inter-

system measurements based on the measured CPICH EcNo of the serving cell according

to the following rules:

• Serving cell parameters Sintrasearch (12dB), Sintersearch (2dB) and SsearchRAT (4dB) are

compared with Squal (CPICH Ec/No – Qqualmin (-18dB)) in S-criteria for cell re-

selection

• UE will measure neighbour cells depending on how parameters are set (if parameters

are not sent UE shall measure all cells)

• 1 - None (Squal > Sintrasearch )

• 2 - WCDMA intra-frequency (Sintersearch < Squal Sintrasearch)

• 3 - WCDMA intra- and inter- frequency, no inter-RAT cells (SsearchRAT < Squal

Sintersearch)

• 4 - WCDMA intra- and inter-frequency and inter-RAT cells (Squal SsearchRAT ) Sintrasearch Sintersearch SsearchRA

T

WCDMA CELL

1 2 3 4

In T-Mobile UK network UE starts

measuring 2G cells when Ec/Io ≤ -

14dB


Cell Reselection 3G -> 2G Procedure

First ranking of all the cells based on CPICH RSCP (WCDMA) and RSSI (GSM)

Rs = CPICH RSCP + Qhyst1(4dB)

Rn= Rxlev(n) - Qoffset1(14dB)

Rn (GSM) > Rs (WCDMA) And

Rxlev (GSM) >QrxlevMin (-111dBm)

Yes No

Cell re-selection to GSM

Neighbour WCDMA or GSM cell calculation with offset

parameter

Serving WCDMA cell calculation, with

hysteresis parameter

UE starts GSM measurements if CPICH Ec/No < qQualMin + sSearchRAT

SintraSearch

SinterSearch

SsearchRAT

CPICH EcNo

qQualMin

Second ranking (R criteria) only for WCDMA cells based on CPICH Ec/No

Rs = CPICH Ec/No + Qhyst2

Rn=CPICH_Ec/No(n)-Qoffset2 Cell re-selection to WCDMA cell of highest

R value


Cell Reselection 3G -> 2G

• Optimum setting of 3G>2G cell reselection triggers depends on:

• Designed utilisation targets for the 3G network (Time on 3G)

• Desired Call Set-up Success Rate (CSSR)

• Minimising the possibility of ping – pong

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

> -4 -4 to -

6

-6 to -

8

-8 to -

10

-10 to

-12

-12 to

-14

-14 to

-16

-16 to

-18

-18 to

-21

<-21

Ec/No [dB]

[%]

Call Setup status statistics for each Ec/No range

• As long as the Ec/No is >-12…-14dB the CSSR is excellent

qQualMin + sSearchRAT ~ -14dB

• To define optimum re-selection thresholds it is important to understand Ec/Io and RSCP performance profiles for T-Mobile UK network

• Bin sizes important



Dense Urban Area

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

> -60 -60 to -

70

-70 to -

80

-80 to -

90

-90 to -

100

-100 to -

112

-112 to -

115

< -115

Ec/No [dB]

[%]

Call Setup status statistics for each RSCP range

In urban area the mapping has been found as: -14dB Ec/No -> ~-102dBm RSCP


• Due to very different fading conditions, it may be necessary to consider different

parameter sets for 3G -> 2G reselection in different scenarios;

• 3G border

• Outdoor, typical outdoor to dedicated indoor (in case of missing 3G indoor)

• Special indoor cases without dedicated 3G where the UE speed is high (e.g.

tunnels)

• For example in 3G border

coverage environment the

EcNo level can be seen to

drop much faster

compared to RSCP

-20

-18

-16

-14

-12

-10

-8

-6

-4

14:1

4:5

2.8

67

14:1

5:2

0.0

97

14:1

5:5

5.0

97

14:1

6:3

0.0

97

14:1

7:0

5.0

98

14:1

7:4

1.0

99

14:1

8:1

6.0

90

14:1

8:5

2.0

91

14:1

9:2

8.0

93

Time

Ec/N

o (

dB

)

-140

-120

-100

-80

-60

-40

-20

0

RS

CP

(d

Bm

)




Check levels every 5s from serving GSM cell

and best 6 GSM neighbour cells

UE starts WCDMA measurements if Rxlev running average (RLA_C) is below or above

certain threshold: RLA_C Qsearch_I and Qsearch_P (GPRS)

UE can select WCDMA cell if the level of the serving GSM and non-serving GSM cells has been

exceeded by certain offset for a period of 5 s: CPICH RSCP > RLA_C + FDD_Cell_Reselect_Offset

UE will re-select WCDMA cell in case it's quality is acceptable:

CPICH Ec/No Minimum_FDD_Threshold

Compare levels of all GSM cells

to WCDMA neighbour

Check quality of neighbour

WCDMA cells, no priorities between

WCDMA neighbours 0

5.0

8:T

his

ma

y t

ake

up

to 3

0s


• Re-selection measurements are controlled by parameter threshold to search WCDMA

RAN cells (QSRI)

• This parameter defines a threshold and also indicates whether these measurements are performed when RLA_C (a running average of received signal level) of the

serving GSM cell is below or above the threshold

• In GSM the UE is usually set to measure the 3G neighbours all the time i.e. Qsearch_I and Qsearch_P are both set to 7


UE starts WCDMA measurements if Rxlev running average (RLA_C) is below or above

certain threshold: RLA_C Qsearch_I and Qsearch_P (GPRS)


• For the the camping in indoor environment the set-up could be :

• Indoor GSM / Outdoor GSM (serving indoor)-> Indoor WCDMA / Outdoor

WCDMA (serving indoor)

• Mobile station measuring WCDMA neighbor only when it is well inside the

building using parameter Threshold to search WCDMA RAN Cells

• The defined set-up can be also used in outdoor environment to push the UEs to

3G as soon as possible from the 2G cell to the border 3G cell



• As a general rule the value for FDD_Qmin parameter can be set to –11…-12 dB

(i.e. for the case where the QqualMin +Ssearch_RAT = -14dB)

How to avoid ping-pong ?

UE will re-select WCDMA cell in case it's quality is acceptable:

CPICH Ec/No Minimum_FDD_Threshold

• The “rule” to set the FDD_Qmin value has not been possible to be fulfilled until the specification change (05.08 v8.18.0, 2003-8) has been implemented to the UEs – as below

QqualMin = -18dB

QqualMin + Ssearch_RAT = -14dB

FDD_Qmin >=-12

Camping in 3G Camping in 2G Camping in 3G

CPICH Ec/No

t

FDD_Qmin >= QqualMin + Ssearch_RAT

Fdd_Qmin mapping

Aif parameter 0 1 2 3 4 5 6 7

Fdd_Qmin (old) [dB] -20 -19 -18 -17 -16 -15 -14 -13

Fdd_Qmin (new) [dB] -20 -6 -18 -8 -16 -10 -14 -12



• 3G <> 2G Cell Re-selection

• 3G<>2G Neighbour Plan Design Guidelines



• 3G ISHO Service Optimisation (AMR and PS Data)


Inter-System Neighbours – Design Guidelines

• Principles for 3G2G neighbour relations

• Neighbouring plan should be kept as simple as possible

• The best neighbours at each point of the 3G coverage border should be in

the list of adjacencies.

• If a 2G layer has a strong interference situation (tight frequency re-use),

layers with less interference could be preferred

• If a 2G layer has high blocking probability, layers with less traffic could be

preferred

• Avoid 2G neighbour lists containing multiple instances of the same RF carrier

• Ensure 2G neighbour plan maximises reuse between cells on same BCCH-

BSIC combination

• Principles for 2G3G neighbour relations

• The overlapping 3G cell should be in the list of adjacencies of all underlying

2G cells

• If the list of overlapping cells includes more than 32 GSM cells, it has to be

shortened

• One solution is to restrict neighbour list to ONE GSM band


Multiple Instances of same RF Carrier in 2G Neighbour

List • Dropped in 3G PS ISHO.

• 9 Compressed Mode Activated, but no HO Command.

• During each phase of CM, RNC selects BCCH1 (2 GSM cells in ncell list BCCH1

BSIC29 & BCCH1 BSIC5) and asks UE to verify BSIC, but never receives MR

with the BSIC reported.

1f SC332 EcNo=-13dB, RSCP=-113dBm.

MC to verified the BSIC.

2G Coverage


3G-2G Adjacencies – Design Guidelines

3G

2G

3G

2G

2G 3G

Co-Azimuthed: N3G = N2G + Cell2G

Co-Sited, Not Co-Azimuthed: N3G = N2GCell1 п N2GCell2 + Cell12G + Cell22G

Not Co-Sited, Not Co-Aziluthed: Manual Design


• 2G->3G: Max # of IS Neighbours per 2G cell (ADJW):

• Max number of 3G neighbours= 32

• However in Nokia implementation to allow the SI2 message in only one BCCH block the 3G neighbours list size should be limited to 10 neighbours.

• 3G->2G: Max # of IS Neighbours per 3G cell (ADJG):

• Max number of 2G neighbours= 32

• Minimize the number of 2G neighbours to improve the performances Nokia suggest to limit the number of 2G neighbours to 16 when possible

• If too many adjacencies are declared the cell will go blocked by system with alarm:

• 7761 RNW O/M SCENARIO FAILURE (BCCH scheduling error) in RN1.5.2ED2

• 7771 WCDMA CELL OUT OF USE (BCCH scheduling error) in RN2.0

• Nokia RNC software Technical Note 46 specifies Restriction on number of cells in

SIB11/12 message for 47 neighboring cells (worst case)

• On one hand the SIB type 11 and 12 messages can contain information on the

maximum of 96 cells (32 intra-frequency cells, 32 inter-frequency cells and 32 GSM

cells), but on the other hand the physical size of SIB data (no more than 3552 bits) has capacity only for 47 neighboring cells.

Inter-System Neighbours – Design Guidelines




• Neighbour planning





Handover Triggering Thresholds set in RNC

Event Triggered Coverage/Capacity based HO fulfilled in RNC

RNC commands the UE to start IS measurements periodically

Measurements are done in Compressed Mode (CM)

UE reports best GSM cells having strongest RSSI results back to RNC

RNC makes HO decision and commands UE to target cell

RSSI measurements and verification for GSM cells

Inter System Handover 3G -> 2G Procedure

• Currently ISHO measurements can be triggered for 5 different reasons

• To measure GSM frequencies, 3G UE needs (a) dual receiver or (b) Compressed Mode capability

• Most UEs currently use CM to ‘create’ gap during which BCCH frequencies on the 2G network can be measured

• CM introduces a delay which depends on the measurement reason and pattern used

• Two types of CM measurements; GSM RSSI & GSM BSIC verification

• Methods for implementing CM are (i) High Layer Scheduling (HLS), (ii) Spreading Factor Halving (SF/2) and (iii) Puncturing

• Handover decision driven by RNC using measurements made by UE

• Compressed Mode affects 3G coverage, capacity and quality

!


Downlink DPCH power UL Quality

deterioration UE Tx power CPICH RSCP CPICH Ec/I0

RAN Internal measurements Configured UE measurements

Initiate Compressed Mode Configure GSM measurements

GSM cell meets HO condition ?

Initiate Handover

Initiate Compressed Mode Configure GSM measurements

UE Reports GSM BSIC measurements

UE Reports GSM RSSI measurements

No

Yes

Is BSCI verification required for PS call ?

No

Yes

Inter System Handover 3G -> 2G Procedure


• BSIC verification always performed for AMR calls – no interrupt in voice call

CN UE Node B

RNC

RRC: Measurement Report

RRC: Measurement Control

NBAP: Radio Link Reconfiguration Prepare

NBAP: Radio Link Reconfiguration Ready

NBAP: Radio Link Reconfiguration Commit

RRC: Physical Channel Reconfiguration

RRC: Physical Channel Reconfiguration Complete

NBAP: Compressed Mode Command






RRC: Handover from UTRAN Command

GSM BSIC Identification

GSM RSSI Measurement

ISHO triggering (5 reasons are possible)

Initial Compressed Mode Configuration

RANAP: Relocation Required

RANAP: Relocation Command

RANAP: IU Release Command

RANAP: IU Release Complete

ISHO 3G -> 2G - AMR Signalling Flow


• In most cases BSIC verification is not required (data interrupt as UE moves to 2G)

• PS makes use of RRC: CELL CHANGE ORDER FROM UTRAN message

UE

Node B

RNC



NBAP: Radio Link Reconf iguration Prepare

NBAP: Radio Link Reconf iguration Ready

NBAP: Radio Link Reconf iguration Commit

RRC: Physical Channel Reconf iguration

RRC: Physical Channel Reconf iguration Complete



RRC: Measurement Control GSM RSSI Measurement



CN

RANAP: SRNS Context Request

RANAP: SRNS Context Response



RRC: Cell Change Order f rom UTRAN

RANAP: SRNS Data Forward Command

ISHO 3G -> 2G - PS Signalling Flow


Compressed Mode

Procedure

• RNC informs UE of CM pattern and 2G neighbour list in RRC:

Measurement Control message

• 3GPP states that the UE must be capable of recording a

minimum number of GSM RSSI measurement samples per

transmission gap

• GSM RSSI measurements are made without acquiring GSM

synchronisation

• UE reports on strongest 6 GSM neighbours at periodic

interval defined by GSMMeasRepInterval

• RNC applies a sliding averaging window to the

measurements provided by the UE

• RNC instructs UE to perform BSIC verification (AMR=always,

PS=as required) using new CM pattern

• BSIC verification needs synchronisation to GSM frame

RNC commands the selected UEs to enter compressed mode and provides compressed mode parameters and neighbour list

During the DL reception gap UE

can make measurements from 2G network

UE measures RSSI of GSM

neighbours and reports these to RNC periodically

RNC makes HO decision

RNC commands UE to decode BSIC of cell with strongest RSSI

RNC sends handover command to UE

WCDMA

IS-HO trigger Target Cell found IS-HO command

RSSI measurements

BSIC verification

TRSSI TBSIC


Compressed Mode Method

• Single frame approach (used in

spreading factor halving and HLS

½ data rate)

• Double frame approach (used in

puncturing and HLS ¾ data rate)

CM Method SF/2

AMR Speech RT Data

SF/2

NRT Data

½ or ¾ Rate HLS

• Compressed mode Methods used in Nokia

!


CMmasterSwitch RNC

Scope Configurable

Yes

Nokia Range

0 (false), 1 (true)

WCEL Yes 0 to 255

WCEL Yes

WCEL Yes

WCEL Yes

MaxNumbUECMcoverHO

PrxTarget

PrxOffset

PtxTarget

PtxOffset WCEL Yes

Default/T-Mobile UK

1

0 to 30 dB

0 to 6 dB

-10 to 50 dBm

0 to 6 dB

16

4 dB

1 dB

40 dBm

1 dB

• Compressed mode can be enabled/disabled on a per RNC basis using CMmasterSwitch

parameter

• The maximum number of UE allowed to be in compressed mode simultaneously can be limited using the MaxNumbUECMcoverHO parameter

• PrxTarget, PrxOffset, PtxTarget and PtxOffset are also used when making the decision

whether or not a UE is allowed to apply compressed mode

• If the cell exceeds Prx/txTarget then one more UE may apply compressed mode during

that radio resource indication period

• If the cell exceeds Prx/txTarget + Prx/txOffset then no more UE may apply compressed

mode during that radio resource indication period

Other Relevant CM Parameters




• 3G<>2G neighbour Plan Verification





• 3G ISHO performance should be analysed from;

• Drive Test Data

• Network Statistics

to obtain an overall picture of ISHO performance

3G ISHO Analysis Overview


• Drive Test Analysis should deliver ISHO performance metrics

such as;

• Number of ISHO attempts

• ISHO & Compressed Mode Success Rate

• Trigger reasons

• Failure Causes

• ISHO Times and Interrupt Delays

• Number of CM cycles needed

ISHO Analysis from Drive Test

Data


ISHO Analysis from Drive Test

Data

Need to analyse each ISHO

attempt to determine;

• Whether it was a necessary ISHO

• What the trigger mechanism was

(helps determine whether trigger

thresholds are set correctly)

• Whether the ISHO was successful

and if not the failure cause (e.g. No

cell found, UE failed BSIC

verification, UE didn’t receive

HandoverFromUTRANCommand)

• ISHO Success Rates

• Process needs to be automated

within the drive test post-

processing tool(s)


PS ISHO Failure Analysis - Example

33%

33%

7%

7%

7%

13%

Network does not isse 'Cell Change Order'

UE responds to 'Cell Change Order' with a 'Failure'

UE does not react to the 'Cell Change Order'

UE does not report any GSM RSSI measurements

UE does not receive the CM 'Measurement Control'

UE fails BSIC verification

• Generating an ISHO Failure Breakdown enables areas for further

optimisation to be identified

• May require further troubleshooting with data logging, RNC counters etc. to

determine root cause of failure

• Analysis should be performed for each service (i.e. AMR, PS etc.)


0%

22%

0%

56%

22%

UE does not read any 2G sys info

UE reads 2G sys info but does not send RACH (no cell re-

sel)

UE reads 2G sys info and sends RACH but does not est.

connection (no cell re-sel)

UE reads 2G sys info and completes 2G cell re-sel but

does not send RACH

UE reads 2G sys info and completes 2G cell re-sel and

sends RACH but does not est. connection

• 78 % of the failures occur after the UE has completed a 2G cell re-selection.

100 % of these 2G cell re-selections was onto a 2G cell which was not in the

3G system neighbour list

• The 2G neighbour lists thus appear to have missing neighbours which

subsequently result in a 2G cell re-selection and a slowing down of the inter-

system handover procedure

• This failure scenario could be used to refine the 3G system inter-RAT

neighbour lists during the post processing and analysis of any drive test data

Example PS ISHO Failure Analysis


GsmMaxMeasPeriod x GsmMeasRepInterval = 12s

ISHO Failure - No Cell Found Failure Example

Tstart = 17:22:41.7

Tstop = 17:22:53.7 MW = 12 s

Compressed Mode started

Compressed Mode stopped


RxLev = -110 + 4 = -106 dBm

4

AdjgRxLevMinHO = -104 dBm

POOR GSM

COVERAGE

No suitable cell

ISHO Failure - No Cell Found Failure Example


ISHO Failure - UE receives

iuv_relocation_prep_fail

• GSM BSIC verification is achieved and

RNC sends iuv_relocation_required

message with the target CI, LAC but

receives an iuv_relocation_prep_fail

message back from the CN

• Failure due to data build error in 2G MSC

• In this failure case no

HandoverFromUTRANcommand is

observed


Analysing ISHO Delay

• ISHO delay ultimately impacts overall ISHO performance > longer the delay,

greater chance of ISHO failing

• Performance impact greater on PS calls (throughput reduction)

• ISHO delay is affected by;

• RNC databuild – 2G neighbour lists provided to UE must be optimum.

The longer the list the longer it takes UE to complete RSSI

measurements

• Radio Plan - Areas of excessive interference will hinder UE’s ability to

decode BSIC and registration procedure

• System design - Implementation of Gs interface enables faster

‘combined’ LAU/RAU registration procedure, rate of broadcasting

system information messages, e.g. SI13

WCDMA

IS-HO trigger Target Cell found IS-HO command

RSSI measurements

BSIC verification

TRSSI TBSIC


ISHO delay, AMR

Ta= activation time

Trssi= RSSI delay time. By default, the network only requires the first RSSI measurements to choose the best GSM target cell. TBSIC= Time for UE to decode the BSIC of the

chosen cell

11.3

2.0

0.2

4.5

0

1

2

3

4

5

Ta

Trssi

Tbsic

Thandover

Total delay

• Measurements taken from Nokia

test network (predominantly

suburban environment) – averaged

over a large number of calls

• BSIC verification process has the

greatest impact on AMR CM delay

• Little scope for reducing this delay

as BSIC verification always required

– but delay in AMR of less

importance as there is no service

gap to user with AMR calls during CM

Thandover

RNC

RRC: Measurement Control (IE tgmp=GSM RSSI)


. .

UE

ISHO triggered: Event 6A/1F/

RRC: Measurement Control (IE tgmp=GSM BSIC)

Handover from UTRAN

. .

. RRC: Measurement report : BSIC decoded

RRC: Measurement report: BSIC no decoded

Ta

Trssi

TBSIC

Total Delay


Ta= activation time

Trssi= RSSI delay time. By default, the network only requires the first RSSI measurements to choose the best GSM target cell.

1.41.2

2.6

0

0.5

1

1.5

2

2.5

3

seco

nd

s

Tactivation

Trssi

Total

ISHO delay, PS

• PS ISHO introduces service affecting

data interrupt (i.e. throughput reduction)

• In most cases PS calls do not require BSIC verification during ISHO process

• BSIC verification required when 2G

neighbour list contains multiple cells on same frequency as the ‘best’ RF carrier

chosen by the RNC

• Need to avoid the need for BSIC verification where ever possible

RRC: Measurement control (IE TGMP=”GSM RSSI”)

RRC: Measurement report

ISHO triggered: Event 6A/1F/

RRC: Cell change Order from UTRAN

Ta

Trssi

RNC UE


Impact of 2G Neighbour List Length

on TRSSI

• UE measures all GSM carriers indicated in the Measurement Control message prior to

sending Measurement Report to RNC

• If UE is not able to measure all GSM carriers within the measurement period it sends an ‘empty’ Measurement Report

• Once UE has measured all 2G carriers (L1) it reports these measurements to L3 which

reports the top 6 strongest to RNC in Measurement Report

• In general GSM RSSI measurement delay increases as the length of the 2G neighbour

list increases. There is however quite a large scatter

0

5

10

15

20

25

30

35

0.0 1.0 2.0 3.0 4.0 5.0 6.0

Delay to the first GSM RSSI Measurement (seconds)

Nu

mb

er

of

GS

M N

eig

hb

ou

rs

• Tests made with Nokia 7600 UE

• The general trend is that the GSM

RSSI measurement delay increases as the length of the 2G neighbour list

increases. There is however quite a

large scatter


• Throughput will be affected by;

• method of CM used (e.g. HLS)

• poor coverage requiring additional re-transmisisons

• Once UE has been instructed to move to 2G throughput drops to zero until UE

has registered (LAU and RAU) on the 2G network

Average = 121 kbps HLS starts

Th

rou

gh

pu

t

Time

Average = 71 kbps

• Example measurements from the

HLS method applied to the PS data

service

• 7 slot transmission gap

• Single frame approach

• 4 frame transmission gap pattern

length

• Throughput reductions

• 64 kbps service: 55 to 15 kbps

• 128 kbps service: 121 to 71 kbps

CM Throughput Reduction – HLS

Example


ISHO Analysis from Network

Statistics

• ISHO counters are found in the Inter-system hard handover measurement table

(M1010Cxx)

• The inter-system handover measurement is carried out in the serving RNC

(SRNC)

• The object of the measurement is the source cell from which the handover is

attempted.

• In case of multiple cells in the AS the object of the measurement is the best cell

of the Active Set during the handover decision (except for

IS_COM_MOD_STA_NOT_POS_(N)RT) where the object is the best active

set cell during the compressed mode preparation phase

• From the measurement point of view the ISHO is divided into two phases:

• Inter-system measurement

• Inter-system handover

• Separate counters for RT and NRT services


Counters for Triggering

Reasons

RNC



If compressed mode can NOT start (before sending Measurement Control to UE) the following counter is incremented:

IS_COM_MOD_STA_NOT_POS_(N)RT

The counter is updated when:

- Admission Control rejects compressed mode request

- Compressed mode can not start due to radio link (or physical channel) reconfiguration failure (BTS or UE reasons)

- ISHO is a parallel procedure (especially NRT)

If compressed mode can start the following counters are incremented:

IS_HHO_W_CMOD_UL_DCH_Q_(N)RT

IS_HHO_W_CMOD_UE_TX_PWR_(N)RT

IS_HHO_W_CMOD_DL_DPCH_(N)RT

IS_HHO_W_CMOD_CPICH_RSCP_(N)RT

IS_HHO_W_CMOD_ CPICH_ECNO_(N)RT

Compressed

Mode start

These counters are not incremented in this phase but they are

incremented After RxLev/BSIC verification phase together with the

ATTEMPT or NO CELL FOUND counters

The counters are updated in the best cell of the active set

CHECK TRAFFIC TABLE COUNTERS

ALSO COUNTERS FOR ISHO START WITHOUT COMPRESSED MODE EXIST (DUAL RECEIVER)

IS_HHO_WO_CMOD_XXX


Counters for measurement procedure failure

If the RNC does not find a suitable cell, one of the following

counters are updated:

IS_HHO_NO_CELL_UL_DCH_Q_(N)RT

IS_HHO_NO_CELL_UE_TX_PWR_(N)RT

IS_HHO_NO_CELL_DL_DPCH_(N)RT

IS_HHO_NO_CELL_CPICH_RSCP_(N)RT

IS_HHO_NO_CELL_CPICH_ECNO_(N)RT

Triggered in the refernce cell when: 1) GsmMaxMeasPeriod measurement reports are received with an rx level

not suitable 2) GsmMaxMeasPeriod measurement reports are received with a BSIC not

verified

RRC: ”Measurement Control”

RNC

RRC: ”Measurement report”



RRC: ”Measurement report” (3,4,5)

These counters are updated when either:

• No suitable GSM target cell is found in terms of RSSI

(RxLev)

• Target cell is suitable (RSSI) but BSIC verification fails

And;

Maximum number of measurement reports have been

received


Compressed Mode Stopped – Case 1

-Stop RxLev measurement -Start BSIC verification phase

-Stop BSIC verification phase (GsmMaxMeasPeriod msg with measurementIdentity = 3 received before the max_meas_interval time window expires)

RNC does receive GsmMaxMeasPeriod ISHO Measurement Reports (meas id = 3) before the above max_meas_interval time window expires.

-GsmMaxMeasPeriod = 12 -GsmMeasRepInterval = 0.5s -max_meas_interval = (12+4)/2 = 8s

13 Measurement Reports: -12 with measurementIdentity = 3 -1 with measurementIdentity =5 (1f -EcNo)

meas_interval < 8 s

IS_HHO_W_CMOD_DL_DPCH_RT = +1 IS_HHO_NO_CELL_DL_DPCH_RT = +1


Counters for ISHO Attempt for RT

If the RNC has found a suitable cell (both RxLev

and BSIC phases), one of the following counters

are updated:

IS_HHO_ATT_UL_DCH_Q_RT

IS_HHO_ATT_UE_TX_PWR_RT

IS_HHO_ATT_DL_DPCH_PWR_RT

IS_HHO_ATT_CPICH_RSCP_RT

IS_HHO_ATT_CPICH_ECNO_RT

The counters are triggered when the RNC sends

the RANAP: RELOCATION REQUIRED

message to the MSC

Only the SRNC can update the counters.

HHO Attempt

Counters


RELOCATION REQUIRED


RNC





MSC X

RxLev meas.

BSIC verif.



Counters for ISHO Attempt for NRT

RNC





If the RNC has found a suitable cell (most

likely only RxLev), one of the following

counters are updated:

IS_HHO_ATT_UL_DCH_Q_NRT

IS_HHO_ATT_UE_TX_PWR_NRT

IS_HHO_ATT_DL_DPCH_PWR_NRT

IS_HHO_ATT_CPICH_RSCP_NRT

IS_HHO_ATT_CPICH_ECNO_NRT

The counters are triggered when the RNC

sends the RRC: CELL CHANGE ORDER

FROM UTRAN message to the UE.

Only the SRNC can update the counters.

HHO Attempt

Counters


CELL CHANGE ORDER X

RxLev meas.

BSIC verification only in case 2 ADJG

with same BCCH


KPI for ISHO measurement procedure

AllcausesAllcauses

Allcauses

RTNxxxCMODWOHHOISRTNxxxCMODWHHOIS

RTNxxxCELLNOHHOIS

RateFailMeasISHO)_(____)_(____

)_(____

___

Compressed

Mode start

No Cell Found

Counters

HHO Attempt

Counters

NO counters

triggering

It’s not possible to distinguish between failures in the RxLev measurement or in the BSIC

verification

It’s not possible to monitor cases like: 1) the UE does not send some measurement reports 2) Drop during measurement 3) parallel procedure (e.g capacity request for NRT)

Max measurement interval not expiring

CM measurement is not interrupted

… measurement fail

… measurement not fail

Accuracy in GSM neighbour planning

Max measurement interval expiring OR

CM measurement is interrupted

KPI is useful for cell level


Call Duration and ISHO procedure triggered

• Monitoring how often ISHO procedure is started ISHO Triggering threshold tuning

• The KPI on cell level does not take into account user mobility. It does make sense on

cluster/RNC level only

• The KPI can help operators in understanding the ISHO strategy

• Same KPI for NRT

(AVG_DCH_HLD_TM_PS_INTER+AVG_DCH_HLD_TM_PS_BACKG at numerator)

• Service level counter at denominator shall be used because allocation duration

counters (Traffic table) incremented in all the cells within Active Set

• Accuracy of the Indicator affected by:

• the number of ISHO procedures which have no ISHO counter incremented (e.g. drop during

CM, see previous slide)

• number of not successful ISHO procedures during the same call due to e.g. wrong ADJ

parameter, rough mobiles etc.

withoutCMwithAllcauses

OS_RT_STA_NOT_PIS_COM_MODD_RTIS_HHO_CMO

ICE/D_TM_CS_VOAVG_RAB_HLcedure_RTr_ISHO_proMinutes_pe

&,

6000

KPI is useful for cluster level and cell level as well


ISHO Attempt UTRAN Failure RT

RNC






Handover Command

When the UTRAN is not able to execute an Inter-System Handover the following counter is triggered:

UTRAN_NOT_ABLE_EXC_ISHHO_RT The counter is triggered when the ISHO fails before the

SRNC sends the handover command to the UE, in the same cell where the ISHO attempt has been updated:

• Relocation Preparation Failure or • TRelocPrep (def. 6s, from Relocation Required to

Relocation Command) expires.

The failure can take place for the following reasons:

• Radio Resource congestion in the target cell • Radio Link setup/addition failure in the BTS (IFHO) • Failure during the Relocation preparation procedure in

the CN (for example ciphering parameter not set properly in 3G MSC, LAC mismatching in RNC/MSC)

• Failure during the Relocation resource allocation procedure in the target BSC

UTRAN Failure

Counter

Relocation Procedure


KPI for ISHO Triggering Reasons

RNC



It’s important to know which is the most frequent triggering reason:

It’s possible to diffentiate between quality and coverage reasons

Understand the network limiting factors:

• CPICH coverage

• Pilot pollution

• UL/DL Service coverage

The triggerning reasons must be enabled: GSMcause…

Allcauses

RTNxxxCMODWHHOIS

RTNxxxCMODWHHOISpercCausexxx

)_(____

)_(______

KPI is useful for cluster level and cell level as well


ISHO UE Failure RT and NRT

RNC






HANDOVER FROM UTRAN

When the UE is not able to execute an Inter-System Handover the following counter is triggered: UE_NOT_ABLE_EXC_ISHHO_(N)RT The counter is triggered when the source RNC receives a failure message from the mobile with the failure cause “configuration unacceptable”. The counter is triggered in the same cell where the ISHO attempt has been updated.

UE Failure

Counter

CELL CHANGE ORDER FROM UTRAN

x

HANDOVER FROM UTRAN FAILURE

CELL CHANGE ORDER FROM UTRAN FAILURE


Unsuccessful ISHO RT and NRT

RNC






Handover Command

When the RNC receives a failure message from the mobile and the cause is not “configuration unacceptable”, one of the following counter is triggered: UNSUCC_IS_HHO_UL_DCH_Q_(N)RT UNSUCC_IS_HHO_EU_TX_PWR_(N)RT UNSUCC_IS_HHO_DL_DPCH_PWR_(N)RT UNSUCC_IS_HHO_CPICH_RSCP_(N)RT UNSUCC_IS_HHO_CPICH_ECNO_(N)RT The counter is triggered in the same cell where the ISHO attempt has been updated. Reason for failure:

• Physical channel failure (the UE is not able to establish – in the target RAT – the phy. Channel indicated in the handover command)

• Protocol error • Inter-Rat protocol error • Unspecified

Handover Failure Cell Change Failure

ISHO Unsuccess

Counters

Cell Change Order (PS)

x


Successful ISHO

RNC






Handover Command

When the source RNC receives the RANAP message “IU RELEASE COMMAND” from the core network, one of the following counter is triggered: SUCC_IS_HHO_UL_DCH_Q_(N)RT SUCC_IS_HHO_EU_TX_PWR_(N)RT SUCC_IS_HHO_DL_DPCH_PWR_(N)RT SUCC_IS_HHO_CPICH_RSCP_(N)RT SUCC_IS_HHO_CPICH_ECNO_(N)RT The counter is triggered in the same cell where the ISHO attempt has been updated. During testing the counter is updated in case of failures (RAB Active failure for RNC internal): for RT use RAB release due to SRNC relocation (relocation not used at the moment)

Handover Complete to BSC

ISHO Success

Counters

IU Release Command

CN



RRC Drop during ISHO RT

RNC






Handover Command

When the source RCC Connection drops during the ISHO, one of the following counter is triggered: CON_DRPS_IS_HHO_UL_DCH_Q_RT CON_DRPS_IS_HHO_EU_TX_PWR_RT CON_DRPS _IS _HHO_DL_DPCH_PWR_RT CON_DRPS _IS _HHO_CPICH_RSCP_RT CON_DRPS _IS _HHO_CPICH_ECNO_RT For RT: TRelocOverall (def. 8s, from Relocation Command to Iu Release Command) expires.

RRC Drop

Counters IU Release Request

CN


RRC Drop during ISHO NRT

RNC






When the source RCC Connection drops during the ISHO, one of the following counter is triggered: CON_DRPS_IS_HHO_UL_DCH_Q_NRT CON_DRPS_IS_HHO_EU_TX_PWR_NRT CON_DRPS _IS _HHO_DL_DPCH_PWR_NRT CON_DRPS _IS _HHO_CPICH_RSCP_NRT CON_DRPS _IS _HHO_CPICH_ECNO_NRT For NRT: RRC-TmrlRCC = T309+ InterRATCellReselTmrOffset expires where: T309 parameter = 5 s (SIB1) InterRATCellReselTmrOffset is hidden parameter 3s

RRC Drop

Counters

IU Release Request CN



KPI for ISHO Failure Rate

RNC






Handover Command

Allcauses

RTxxxATTHHOIS

RTSuccessNbrRTAttOverRateFailISHO

____

__1_____


Allcauses

OS_NRT_STA_NOT_PIS_COM_MODD_NRTIS_HHO_CMO

NRTxxxHHOISSUCC

NRTRateFailISHOOverall

&,

____

1____Cell Change Order (PS)

Nbr_Success_RT =

RAB_ACT_REL_CS_VOICE_SRNC, SRNC Reloc/IFHO off SUM(SUCC_IS_HHO_XXX_RT), SRNC Reloc/IFHO on

Allcauses

Allcauses

NRTxxxATTHHOIS

NRTxxxHHOISSUCC

NRTAttOverRateFailISHO____

____

1_____


OS_RT_STA_NOT_PIS_COM_MODD_RTIS_HHO_CMO

RTSuccessNbrRTRateFailISHOOverall

&,

__1____

KPI are useful for cluster level and cell level as well


ISHO Analysis from Network

Statistics

CN UE Node B RNC



NBAP: Radio Link Reconf iguration Prepare

NBAP: Radio Link Reconf iguration Ready

NBAP: Radio Link Reconf iguration Commit

RRC: Physical Channel Reconf iguration

RRC: Physical Channel Reconf iguration Complete







RRC: Handover f rom UTRAN Command

GSM BSIC Identification

GSM RSSI Measurement



RANAP: Relocation Required

RANAP: Relocation Command




• ISHO success rate on Beckton3 (URKKT03) had always been poor (<80%) with

UTRAN_NOT_ABLE_EXEC_ISHO_RT the largest failure counter (>90% increments on

<6 cells)

ISHO Analysis from Network Statistics UE doesn’t receive HandoverFromUTRAN Command

• Similar pattern was observed across a

number of other RNCs



UTRAN_NOT_ABLE_EXEC_ISHHO_RT counter according to customer documentation


PROCEDURE IS STARTED.....



GSM SUITABLE CELL IS FOUND AND BSIC VERIFICATION IS REQUIRED....

1

BSIC IS CORRECTLY VERIFIED.... 2

RELOCATION IS REQUIRED TO MSC... Timer TRelocPrep (6s) is started

3



Timer TRelocPrep (6s) expires Relocation Cancel msg is sent to MSC.....

Note that, in the meantime, no parallel procedure is allowed.....




• Neighbour planning






3. Decision Algorithm

UE Tx Power (Event 6A) •Threshold: GsmUETxPwrThrXX •L3 filter: GsmUETxPwrFilterCoeff •Hysteresis margin: GsmUETxPwrTimeHyst •Data rate threshold HHOMAxAllowedBitrateUL

UL Quality •Timer: ULQualDetRepThreshold •Data rate threshold HHOMAxAllowedBitrateUL

DL DPCH power •Threshold: GsmDLTxPwrThrXX •Data rate threshold HHOMAxAllowedBitrateDL

(XX=AMR,CS,NrtPS,RtPS)

CPICH RSCP (Event 1F) •Thresholds: HHoRscpThreshold HHoRscpCancel L3 filter: HHoRscpFilterCoefficient •Timers: HHoRscpTimeHysteresis HHoRscpCancelTime

CPICH Ec/Io (Event 1F) •Thresholds: HHoEcNoThreshold HHoEcNoCancel •L3 filter: EcNofilterCoefficient •Timers: HHoEcNoTimeHysteresis HHoEcNoCancelTime

AdjgTxPwrMaxTCH AdjgRxLevMinHO (n) GsmMeasAveWindow

GsmMeasRepInterval GsmNcellSearchPeriod GsmMinMeasInterval GsmMaxMeasPeriod

Handover Execution

2G-to-3G back prevention

GsmMinHoInterval

2. GSM measurement reporting

Inter System Handover 3G -> 2G Tuning

1. Triggering

2. GSM measuring

3. Decision


Inter System Handover 3G -> 2G Tuning

The IS-HO process consists of several independent sub-processes. Therefore the optimisation can be applied to each sub-process individually:

1. Triggering process:

• Parameters that belong to this process defines the starting of the GSM measurements: filters, hysteresis, timers and thresholds

2. GSM Measurement reporting process

• Following parameters control the reporting of the GSM measurements

• GsmMinMeasInterval: Establish minimum time between successive GSM measurements

• GsmMaxMeasPeriod: Maximum duration of the GSM measurements in CM

• GsmMeasRepInterval: Reporting period of the GSM measurements during CM

3. Decision process:

• Parameters that participate in the selection of the best target cell:

• AdjgRxLevMinHO(n): Minimum RX level of the GSM cell to do handover

4. ISHO cancellation parameters:

• Cancellation parameters are built for CPICH EcNO and CPICH RSCP triggering functionality only


• Each triggering procedure makes use of filters, hysteresis and thresholds which

are used to control the inter-system handover behaviour

• The purpose of the hysteresis and filters is to improve the accuracy of the

measurements

• The purpose of the thresholds is to control 3G boundary of the different

services

• Each 3G cell had in average 4 GSM neighbour cells.

• By modifying the network configuration (blocking of cells etc), the radio conditions

in each route was aconditioned so that the need for an inter-system handover

was due to the triggering condition studied

• For example, when the UE Tx power trigger was studied, the radio

conditions along the routes were modified so that the uplink path was the

critical one, in this way the IS-HO performance was not affected by other

reasons e.g. poor CPICH EcNo

• During the tests the following traces were recorded:

-The coordinates where the CM starts

-Unsuccessful/successful ISHO events

ISHO Triggering Process


Different CPICH EcNo thresholds were used with setting 3:

IS-HO performance at threshold values of –11 dB and –12 dB is very similar.

Routes EcNo (IS-HO Success) % IS-HO

Threshold /(Attempts) success rate

Route 2 -11 12/12 100%

-12 12/12 100%

-14 3/9 33.3%

Route 4 -11 7/9 77.7%

-12 8/9 88.8%

-14 7/9 77.7%

Route 6 -11 13/16 81.25%

-12 13/16 81.25%

-14 9/20 45%

-11 32/37 86.50%

Total -12 33/37 89.20%

-14 19/38 50%

ISHO Triggering Process – Ec/Io Threshold


Routes RSCP (IS-HO Success) % IS-HO


-105 9/9 100%

Route 2 -106 9/9 100%

-107 8/9 88%

Route 3 -105 10/10 100%

-106 14/15 93.3%

-107 11/15 73.3%

-105 8/9 88%

Route 4 -106 8/9 88%

-107 6/9 66%

-105 21/25 84%

Route 6 -106 16/20 80%

-107 -- --

-105 48/53 90.5%

Total -106 47/53 88.7%

-107 25/33 75.8%

IS-HO performance at threshold values of –105 dB and –106 dB is very similar.

ISHO Triggering Process – RSCP Threshold


Routes UE Tx Pwr (IS-HO Success) % IS-HO

Threshold /(Attempts) Success rate

-1 6/6 100%

Route 2 -3 6/6 100%

-5 6/6 100%

-1 10/10 100%

Route 3 -3 9/9 100%

-5 9/10 90%

Route 4 -1 6/6 100%

-3 6/6 100%

-5 6/6 100%

-1 22/30 73.30%

Route 6 -3 27/30 90.00%

-5 22/25 88%

-1 44/52 84.60%

Total -3 48/51 94.00%

-5 43/47 91.50%

The results show that a threshold= -3 dB is large enough to provide a good performance.

ISHO Triggering Process – UE Tx Power Threshold


Routes RSCP (IS-HO Success) % IS-HO


-1 0/6 0%

Route 2 -2 0/6 0%

-3 6/6 100%

-1 1/6 16.6%

Route 4 -2 1/6 16.6%

-3 6/6 100%

-1 6/20 --

Route 6 -2 -- --

-3 13/20 65%

-1 7/32 21.9%

Total -2 1/12 8.3%

-3 25/32 78%

Clearly, the IS-HO performance at the threshold GsmDlTxPrwAMR =–3 dB is the best: excellent performance (100 % success rate) at medium speeds but degradation occurs at high speed (route 6).

ISHO Triggering Process – DL Tx Power Threshold


Parameter Suggested values

HHoEcNoThreshold -12 dB

HHoEcNoCancel -9 dB

CPICH EcNo HHoEcNoCancelTime 640 ms

HHoEcNoTimeHysteresis 100 ms

EcNoFilterCoefficient 600 ms

HHoRscpThreshold -105 dBm

CPICH RSCP HHoRscpCancel -103 dBm

HHoRscpFilterCoefficient 200 ms

HHoRscpTimeHysteresis 100 ms

HHoRscpCancelTime 640 ms

UL Tx Power GsmUETxPwrThrAMR -3 dB

GsmUETxPwrFilterCoeff 10ms

GsmUETxPwrTimeHyst 320ms

DL DCH GsmDLTxPwrThrAMR -3 dB

UL Quality ULQualDetRepThreshold 0.5s

Measurement GsmMeasRepInterval 0.5 s

Reporting GsmNcellSearchPeriod 0

Parameters GsmMaxMeasPeriod 20 meas. report

GsmMinMeasInterval 2s

HO AdjgRxLevMinHO -98 to -100 dBm

Decision AdjgTxPwrMaxTCH 33 dBm

Algorithm GsmMeasAveWindow 6 meas. report

ISHO Triggering Process – Parameter Summary


GSM Measurement Reporting Process

RNC



GsmMeasRepInterval (default 0.5s)

GSMMaxMeasPeriod Max 6 GSM cells reported




Handover Command

Handover Complete

Handover Failure

Triggering

Details about the measurements

• Inter-system measurement stops if RNC has not been able to perform inter-system handover after GSMMaxMeasPeriod (value 10, default =20, 1…20, step 1 meas report)

• RNC could not initiate inter-system measurements if:

• The UE has « recently » performed an inter-system HO: GSMMinHoInterval / 10s, 0…60, step 1s

• An inter-system HO « recently » fails for this UE: GSMMinMeasInterval / 10s, 0…60, step 1s

The first measurement report has info from the best GSM cell: BCCH freq & RSSI, no filtering

used in UE


• The RNC databuild parameters GsmMeasRepInterval (0.5 s) and

GsmMaxMeasPeriod (10 measurement reports) define the maximum combined

time which may be used for GSM RSSI measurements and BSIC verifcation, i.e.

5 secs

• If GSM RSSI measurements are completed successfully and if there is only a

single GSM neighbour on the strongest GSM RF carrier then the RNC issues the

Cell Change Order from UTRAN message

• If GSM RSSI measurements are completed successfully and if there are multiple

GSM neighbours on the strongest GSM RF carrier then the RNC instructs BSIC

verification only for the neighbours on that RF carrier

• Both RSSI measurements and BSIC verification make use of a 7 slot

transmission gap every 4 radio frames



GsmMeasRepInterval (default value=0.5 seconds)

• The GSM measurement reporting interval given by this parameter should be

kept to 0.5 seconds (default value)

• Increasing the reporting interval would increase the IS-HO process delay

• Besides, accuracy requirements related to the GSM measurements in

compressed mode are given for a reporting interval of 0.5 seconds (480ms TS

25.133)

GsmMaxMeasPeriod ( default value = 20 measurement reports)

• This parameter controls the maximum compressed mode duration time for

each GSM RSSI and BSIC decoding measurement process. The duration of

this parameters in seconds is given by:

max_meas_time (s) = GsmMaxMeasPeriod x GsmMeasRepInterval

• Thus, the default value of GsmMaxMeasPeriod in seconds is 12



GsmMinMeasInterval (default value=10 seconds):

• In case of an unsuccessful IS-HO attempt, the network will deactivate compressed

mode for a time period given by this parameter value

• The network will reactivate automatically compressed mode after the timer has

expired unless a cancellation event is sent by the terminal during that period

• In case IS-HO cancellation event does not occur, and the mobile is leaving clearly the

3G boundary, then there is a death time equal to 10 seconds in which the UE is not

able to attempt new GSM measurements in case of an unsuccessful IS-HO.

• Experience in the field has shown that the probability of having a cancellation event

(event 1B,1E), after an IS-HO was requested, is very low and therefore the

GsmMinMeasInterval is reduced to a lower value (2 seconds) to speed up the

reactivation of the compressed mode

Trigger threshold

Cancelation threshold

Terminal in Compressed mode

Unsuccessful ISHO 8 s 10 s

Terminal in normal mode

Terminal in Compressed mode

Compressed mode is reactivated automatically by

the network af ter the GsmMinMeasInterval timer

expires



• The Inter-system handover decision process need to be optimized so that the network selects the best GSM target cell which compliance with the following formula:

AVE_RXLEV_NCELL(n) > AdjgRxLevMinHO (n) + max( 0, AdjgTxPwrMaxTCH (n) -

P_MAX )

• AVE_RXLEV_NCELL(n): is the average RSSI level from the cell n

• P_MAX is the maximum power in GSM classmark (+33 dBm)

• AdjgTxPwrMaxTCH (n): is the maximum power of the traffic channel in GSM (default=+33 dBm)

• AdjgRxLevMinHO (n): Minimum required RX level of the target GSM cell to do handover (default=-95 dBm)

• Thus, if P_MAX and AdjgTxPwrMaxTCH (n) default values are used, then the formula simplifies to:

AVE_RXLEV_NCELL(n) > AdjgRxLevMinHO (n)

• AdjgRxLevMinHO(n): The sensitivity of a GSM MS is –104/-102 dBm in GSM 900 MS and –100/-102 dBm for DCS 1800 MS (TS 45.005)

• Since the handover decision process usually only uses one RSSI measurement sample per neighbour with an accuracy of +/- 6 dB (from specs), it may happen that the terminal reports a GSM level of –95 dBm when actual level is –101 dBm

• Could consider using average at RNC with lower threshold value

Handover Decision Process Optimisation


Optimized values

CPICH EcNo

HHoEcNoCancel 3dB above EcNo threshold

HHoEcNoCancelTime 640 ms

CPICH RSCP

HHoRscpCancel 3dB above RSCP threshold

HHoRscpCancelTime 640 ms

• The cancellation thresholds 3dB above the triggering works the best

• Values of the cancellation filters (HHoEcNoCancelTime and HHoRscpCancelTime) depend actually on coverage i.e. in case of fast changes then the filters values should be relatively long to avoid possible ping-pong (initiate IS-HO, cancel, initiate and so on…)

• In case some such ping-pong noted then longer cancellation filters should be tried

ISHO Cancellation Parameter Optimisation


3G->2G for PS data (network controlled cell reselection)

•Uses compressed mode (in both RAN & mobile) for measuring how good 2G coverage exists, before RNC moves the control of the call to BSC

•No resources reserved in BSS in advance; thus the PS data call continues on best effort basis, like all PS data calls in 2G

•No need for BSIC identification

2G->3G for PS data (UE controlled cell-reselection)

•Terminal measures neighboring cells during different time slots (no compressed mode needed) for measuring how good 3G coverage exists, before BSC moves the control of the call to RNC •No resources reserved in RNC in advance; thus the PS data call continues on best effort basis in 3G

WCDMA <-> GSM PS data inter-working functionality


• Throughput degradation in the 3G->GSM frontier exists partly due to the CM activation but also due

to the poor 3G radio conditions

• Below is the average throughput in good and poor 3G coverage

• The definition of poor and good coverage is defined as :

• Good 3G coverage: CPICH EcNo>=-10 dB, CPICH RSCP>-100 dBm

• Poor 3G coverage: CPICH EcNo<-10 dB, CPICH RSCP< -100 dBm

ftp Throughput in ISHO Zones, 384 RAB (server

x.x.x.111 /121)

0

5

10

15

20

25

30

35

40

45

50

1

kB

ps

Good 3G coverage

Poor 3G coverage

GPRS


x.x.x.111 /121)

0

1

2

3

4

5

6

7

1

kB

ps

Good 3G coverage

Poor 3G coverage

GPRS


x.x.x.111 /121)

0

2

4

6

8

10

12

14

16

18

1

kB

ps

Good 3G coverage

Poor 3G coverage

GPRS

6.51

5.41

2.7

17 %, downgrade

43.03

37.76

2.7

12 %, downgrade

15.35

13.82

2.7

10 %, downgrade

10 to 17 % Service downgrade due to coverage and CM between good and poor 3G coverage areas

Throughput in poor coverage includes throughput during compressed mode .

64k RAB 128k RAB 384k RAB

WCDMA <-> GSM PS data inter-working functionality


Cell Change Order (CCO) to GSM

• Triggers

• GsmUETxPwrThrNrtPS (range -10..0 dB,default: -1 dB): This parameter determines the UE

TX power threshold for a non-real time PS data connection. Values –1 dB and –3dB tested.

• GsmDLTxPwrThrNrtPS (range –10..0 dB,default;- 1 dB): This parameter determines the

downlink DPCH TX power threshold for a non-real time PS data connection. Values –1dB

and –3 dB tested.

• Other Parameters Parameter Default Values used

HHoEcNoThreshold -22 dB -12 dB

HHoEcNoCancel - 20 dB -9 dB

CPICH EcNo HHoEcNoCancelTime 1280 ms 640 ms

HHoEcNoTimeHysteresis 640 ms 100 ms

EcNoFilterCoefficient 600 ms 600 ms

HHoRscpThreshold -105 dBm -105 dBm

CPICH RSCP HHoRscpCancel -100 dBm -103 dBm

HHoRscpFilterCoefficient 200 ms 200 ms

HHoRscpTimeHysteresis 640 ms 100 ms

HHoRscpCancelTime 1280 ms 640 ms

UL Quality ULQualDetRepThreshold 5 s 0.5s

UL Tx Power GsmUETxPwrThrNrtPS -1 dB -1, -3 dB

GsmUETxPwrFilterCoeff 10 ms 10ms

GsmUETxPwrTimeHyst 1280 ms 320ms

DL DCH GsmDLTxPwrThrNrtPS -1 dB -1, -3 dB


• Required equipment for 3G/2G neighbour refining

• 2G and 3G scanners + UE in dual mode (outdoor measurements)

• 3G->2G neighbours could be optimized with a prioritization algorithm.

• Priority 1: ISHO failure

• Find the missing neighbours which cause the call failure during the drive test.

• Careful analysis with 2G Scanner and UE data

• Priority 2: poor 3G coverage but good GSM coverage

• 2G and 3G scanner data used to compare signals.

• 3G signal is compared to ISHO triggering parameter values (e.g.

HHoRSCPThreshold, HHoEcNoThreshold)

• GSM neighbor is added if the ISHO condition is met. (2G RSSI>

GsmncellRxLevMinHO)

• Priority 3: good 3G and GSM coverage (probable ISHO when going inside the

building

• 2G and 3G scanner data used to compare signals.

• GSM neighbours of this class are added only when there is still room to reach

the max # of ADJG per cell.

• GSM RSSI > GSMncellRXLevMinHO +Indoor loss (15-25 dB)

3G/2G Neighbour Verification Process - Example


Module 5 – Inter System Handover Optimisation

Summary

• Proper Cell reselection will improve call setup success, it

can be managed with parameters, some margin should be

left to avoid ping-pong

• Good inter-system neighbour planning is key to maintain

the service, the neighbour list should be not too long

• Compressed mode delays the ISHO, for PS data it is

shorter due to no need of BSIC decoding but PS

throughput will be degraded during CM


Version control

Version Date Status Owner

RAN04 2006 Base version Mike Roche, Steve Hunt,

Gareth Davies, Pekka Ranta

RANOP1_5 - ISHO.pdf

Documents

g cells

g gsm target cell

g procedure

g ranop

g coverageinterference

serving cell

g isho analysis

wcdma cell calculation