3G OPtimizer 3.0 for Celcom

3G Opt ptim imize izerr sup s uppor portt for f or Celc lcom om – Malaysi laysia a Prepared Prepared By : Ab A b h i s h ek Rath Rat h Ranj Ra njan an Padh Padhii Reviewed Reviewed B y Caecilia Ca ecilia Ad ji

Proj roje ect Info Inform rma ati tion on Customer Information Frequency Band Vendor Equipment Activity Areas RNCs involved Activity period period License and Build Details :

: Celcom, Malaysia : 3G : NSN : Kepong and SbgHiTech SbgHiTech : 7 RNCs : 12th Sep 2011 till 17 th Feb 2012

Proj roje ect Info Inform rma ati tion on Customer Information Frequency Band Vendor Equipment Activity Areas RNCs involved Activity period period License and Build Details :

: Celcom, Malaysia : 3G : NSN : Kepong and SbgHiTech SbgHiTech : 7 RNCs : 12th Sep 2011 till 17 th Feb 2012

Executive Summary of the Optimizer Optimizer project delivery for Celcom, Celcom, it was was done in 2 phases phases • As part of for the below 2 regions respectively.

Ø Phase 1 : Kepong RNCs Ø Phase 2 : SbgHiTech RNCs

Summary of Acti vit ies fo r the Opti Opti mizer Tool Tool Usage

• Site and Antenna database update • Neighbour Audit Audit & Optimization • Scrambling Code Analysis Analysis and Collision Detection • Interference Analysis • Capacity Analysis

Opt ptim imize izerr Targ Target et Area Ar ea – Kepong & SbgHiTe SbgHiTech ch

Kepong

SbgHiTech

RNC-KPR20 RNC-KPR24

RNC-SHR11 RNC-KPR19 RNC-SHR18 RNC-KPR17 RNC-SHR21

Sit ite e and Ant A ntenna enna database database update upd ate • As the 1st step of the Optimizer activity, the Site & Antenna database was updated based on the input provided by Celcom.

• The site database update was done for : Ø 3G data for all 7 RNCs of Kepong and SbgHiTech Ø 2G/3G data for other vendors in central region

• Subsequently the updates were done on regular intervals to have the most recent data available in the tool. NOTE

:

• During the first phase of the adjacency optimization for Kepong , it was discovered that there were discrepancies discrepancies with with the site database provided by Celcom.

• It was re-vetted by Celcom and another set of updated database was provided which was then used for the subsequent activities.

Neighbour Audit & Optimization

Neighbour Audit & Optimization – Deletion • As part of the neighbor optimization ,the unwanted neighbors were deleted from the network . • The deletions were done based on : Ø Distance (faraway neighbors) Ø HO KPI (neighbors with poor HO KPIs)

• The types of adjacencies considered : Ø ADJS (same frequency) Ø ADJI (inter frequency) Ø ADJG (3G to 2G)

• The criteria followed for the deletion activities are as follows. Type of nbr optimization

ADJS criteria

Faraway nbr deletion HO KPI based nbr deletion

ADJI criteria

ADJG criteria

Distance > 5 Km SHO Attemp ts < 5 AND SHO Success Rate % < 50 %

IFHO Attemp ts < 10 AND IFHO Success Rate % < 50 %

ISHO Attempts < 5 AND ISHO Succ ess Rat e % < 20 %

Neighbour Audit & Optimization – Deletion Summary of Deletion activities Deletion cri teria

RNCs targeted

Date of submissio n to Celcom

Date of implementation

Total no of nbr deleted

Distanc e

KPRN20

21st Oct 2011

24th Oct 2011

156

Distanc e

KPRN17&19

31st Oct 2011

2nd Nov 2011

204

Distanc e

KPRN24

3rd Nov 2011

4th Nov 2011

60

Poor HO KPI

KPRN17&24

17th Nov 2011

9th Dec 2011

620

Poor HO KPI

KPRN19&24

22nd Nov 2011

9th Dec 2011

438

Poor HO KPI & Distanc e

SHRN11,18&21

3rd Jan 2012

9th Jan 2012

2753

• Each nbr delete recommendation was checked individually by the tools team before implementation.

• Every time the evaluation was done in 3 phases. • Post implementation KPI monitoring done by local Optimization team Optimizer plan

Evaluated by too ls team

Evaluated by local NSN optimi zation team

KPI monitoring by local NSN optimization team

Evaluated by Celcom

Implemented

Neighbour Audit & Optimization – Deletion Based on Distance (Far Away Neighbours) – Sample snapshots Kepong RNCs

Kepong RNCs

IFHO Attempts [N] , IFHO Success Ratio [%]

Kepong RNCs

SbgHiTech RNCs

Neighbour Audit & Optimization – Deletion Based on K PI ( ADJ S) – Sampl e snapshot s Kepong RNCs

Kepong RNCs

SbgHiTech RNCs

SbgHiTech RNCs

Neighbour Audit & Optimization – Deletion Based on K PI ( ADJ I) – Sampl e snapshot s

Kepong RNCs

Kepong RNCs

Kepong RNCs

SbgHiTech RNCs

Neighbour Audit & Optimization – Deletion Based on K PI ( ADJ G) – Sample snapshots

Kepong RNCs

SbgHiTech RNCs

SbgHiTech RNCs Kepong RNCs

Neighbour Audit & Optimization – Stats Monit oring Sample KPI moni tori ng pos t imp lementation of nbr delete

Neighbour Audit & Optimization – Deletion • Nbr delete plans were implemented by NSN with Celcom’s approval • Every time after the implementation , the consistency checking was performed by the tools team to ensure the successful implementation of the plan. • Post implementation detailed KPI monitoring was done by the local optimization team. • It was obs erved that no KPI degradation happened after each impl ementation & the same was acknowledged by Celcom .

Neighbour Audit & Optimization – Creation • In Optimizer Neighbours were created Automatically based on the used defined Thresholds collected from measurements and distance. • The below Parameters in Optimizer for the specific Adjacency type. ADJS Creati on Criteri a

Kepong RNCs nbr add Adj acen cy Type

KPI criteria

ADJS

SHO Attemp ts > 5 SHO SR > 80%

ADJI

IFHO Ec/No > -15 dB IFHO RSCP > -95 dBm

ADJG

X

SbgHiTech RNCs nbr add

Same site missing nbr

1st tier missing nbr

√

(checked upto 4 kms)

√

√ √

X


X

KPI criteria SHO Attemp ts > 5 SHO SR > 70% IFHO Ec/No > -15 dB IFHO RSCP > -90 dBm ISHO Attempt s > 50 ISHO SR > 50%

Same site missing nbr

1st tier missing nbr √

√

(checked upto 2 kms) √

√

(checked upto 2 kms) √

√


• Each of the recommended nbr from Optim izer was verifi ed using the Map View of Optimizer.

Neighbour Audit & Optimization – Creation • While creating the missing nbrs, apart from KPI criteria, the following points were considered religiously . Ø Same site missing nbrs are created Ø 1st tier / very nearby nbrs are created Ø Maximum nbr list length rules considered Ø Celcom’s multilayer nbr strategy rules followed Maximum n br lis t length rules

Ø It was ensured that ,with the new addition of

the respective nbrs fall in the given list length rules. Ø This was done because in past Celcom had

issues with the cells when nbrs were added crossing the given list length.

Rules : Ø ADJS <= 31 Ø ADJI <= 31 Ø ADJG <= 31 Ø ADJx <= 47

Neighbour Audit & Optimization – Creation Celco m’s 3rd Carrier Mobi lit y, Neighbors Strategy

Inter BTS

Intra BTS

ADJS

F3

F3

ADJI

ADJS

F2

F2 ADJI ADJI

F1

F1 ADJS

ADJG

ADJG

GSM

GSM

Intra BTS

Inter BTS

ADJI : F1<- F2, F2<-F3, F1<-F3, ADJS : cell s with in s ame car rier

• ADJI : F3 F2, F3->F1*, F2 F1 • ADJS : cel ls with in s ame c arrier

ADJG : F1->GSM

* Ap pl icab le to site w /o 2 nd carrier

Neighbour Audit & Optimization – Creation

Kepong

Neighbo ur additi on for Kepong

• During the evaluation of the Optimizer’s 1 st phase of nbr addition recommendation by the local optimization team for Kepong, it was highlighted that around 900+ 1 st tier nbrs along with same site missing nbrs creation were missing in the Optimizer recommended list. • It was agreed and the Optimizer tools team re-worked on these requirements and came up with approximately additional 3000+ ADJS and ADJI over to the highlighted 900+ missing 1tier missing nbr to meet the local team’s nbr add strategy. Summary of Creation activities for Kepong Creation criteria

RNCs targeted



Total no of nbrs created

Distance & Measur ement

All KPRN RNCs

6th Dec 2011

9th Dec 2011

5808

• Here is the break up of the total nbr addition:

Adj acen cy type

Cou nt

ADJS

5201

ADJI

607 Total

5808


Kepong

Some Exampl es of Very Near Missing ADJS Creations for Kepong

Missing Neighbo ur at 330 Mts

Nearby and Very High At tempt ( > 25000 ) Neighbo ur Missi ng


Kepong

Neighbo ur additio n fo r Kepong - Implementation

• Since not all part of the nbr recommendation was done using Optimizer, the XML plan was generated manually outside Optimizer. • The plan was implemented by NSN with Celcom’s approval on 9 th Dec 2011. • Post implementation KPI monitoring was done by NSN. Neighbo ur addition fo r Kepong – KPI monitori ng

• The nbr addition for all 4 Kepong RNCs showed very good improvement . Outline of the monitori ng : Ø Weekly CS Voice DR comparison showed 7.8% improvement by

comparing 2-8/Dec against 9-15/Dec. Ø Daily CS Voice DR of 15/Dec against 8/Dec showed improvement of

5.5%.

Neighbour Audit & Optimization – Creation Neighbo ur addition fo r Kepong – KPI monitori ng

Kepong


SbgHiTec h

Neighbour addition for SbgHiTech

• The 2nd batch of the nbr add recommendation for SbgHiTech region was solely out of Optimizer. • No manual work was done for additional recommendation over to the ones from Optimizer. Summary of Creation activities for SbgHiTech Creation criteria

RNCs targeted



Total no of nbrs created

Distance & Measur ement

All KPRN RNCs

1st Feb 2012

3rd Feb 2012

11269

• Here is the break up of the total nbr addition:

Adjac enc y t ype

Cou nt

ADJS

5260

ADJI

3133

ADJG

2876

Total

11269

Neighbour Audit & Optimization – Creation Some Exampl es of Very Near Missing ADJS Creations for SbgHiTech

Missing Neighbour at less th an 200 Mts

SbgHiTec h


SbgHiTec h

Neighbou r addi tio n for SbgHiTech - Implementation

• Since Optimizer was solely used for making recommendation for SbgHiTech, the Optimizer plan was used for implementation. • The plan was implemented by NSN with Celcom’s approval on 3 rd Feb 2012. • Post implementation KPI monitoring was done by NSN. Neighbo ur additi on f or SbgHiTech – KPI monitori ng Drop Call Rate & Total CS Traffic for SHR11,18,21 Combined

• No remarkable improvement 0.70

60,000.00

Nbr Add 0.60

0.50

e t a R l 0.40 l a C p 0.30 o r D

obs erved in DCR for Sbg HiTech

0.59 50,000.00

0.55 0.48

0.52

0.50

0.47

0.47 0.44

0.43 0.43

0.40 0.40 0.40

0.51

0.49

0.46

0.45 0.46

0.43 0.43

0.20

RNCs after major nbr addition.

• The KPIs are still being moni tored due to a long weekend falling r ight after the

10,000.00

0.10

0.00

C I F F A R T 30,000.00 S C L A 20,000.00 T O T 40,000.00

implementation. F r i 0 1 .2 7 .2 0 1 2

S a t 0 1 .2 8 .2 0 1 2

S u n 0 1 .2 9 .2 0 1 2

M o n 0 1 . 3 0 .2 0 1 2

T u e 0 1 . 3 1 .2 0 1 2

W e d 0 2 . 0 1 .2 0 1 2

T h u 0 2 . 0 2 .2 0 1 2

F r i 0 2 . 0 3 .2 0 1 2

S a t 0 2 . 0 4 .2 0 1 2

S u n 0 2 . 0 5 .2 0 1 2

M o n 0 2 . 0 6 .2 0 1 2

T u e 0 2 . 0 7 .2 0 1 2

W e d 0 2 . 0 8 .2 0 1 2

T h u 0 2 . 0 9 .2 0 1 2

F r i 0 2 .1 0 .2 0 1 2

S a t 0 2 .1 1 .2 0 1 2

S u n 0 2 .1 2 .2 0 1 2

M o n 0 2 .1 3 .2 0 1 2

T u e 0 2 .1 4 .2 0 1 2

W e d 0 2 .1 5 .2 0 1 2

0.00

Scrambling Code Analysis and Colli sion Detection

Scrambling Code Analysis and Collision Detection • As per the original plan, Optimizer was supposed to be used for SC collision detection and correction. • According to Celcom SC plan strategy ,it has to follow grouping pattern while making the SC correction. • It was found that , the tool woul d not be able to accommodate this strategy while making this correction.

• Later it was mutually agreed that Optimizer would be used for detecting the SC collisio n only and it would generate the collision list to be shared

with the local optimization team. • Based on the list, the opti mization team would use their in-house macro to come up with the SC correction list manually.

Scrambling Code Analysis and Collision Detection • Celcom follow a typical SC pattern to easily plan the SC for newly added sites. • As per Celcom’s SC planning strategy, the whole network has been distributed in hexagonal clusters. Each hexagon contains certain no of sites falling in it geographically & each hexagon is allocated with a set of SC group (R2-R8).Any site falling in the specific hexagon, would have to use the SC from the list of codes available Hexagonal c luster dis tribution for SC Groupin g (R1-R8)

in that particular group.

Scrambling Code Analysis and Collision Detection • After selecting the SC group, each cell pertaining to that site would select the codes from the Code Group (CG1 – CG63) for allocation. • The SC allocation for cells in a particular site is with an increment of 8 codes . • e.g. cellA is allocated with SC X, then cellB would be X+8 and cellC would be X+8+8. Note :

• All INBUILDING sites would use SC group “R1”. • All Macro sites would use rest of the SC group i.e. R2-R8 . Code Group allocation pl an CG1 - CG63

• CG52-63 are reserved for special events.

Scrambling Code Analysis and Collision Detection

Kepong

• SC collision detection was done for all Kepong RNCs. • 1749 collisi ons were detected in total. • Based on this list, the NSN optimization team used their in-house macro and came up with the SC collision correction plan. • Total no of SC changes : 1197 Sample SC collisio n path f rom Optimizer


Kepong

Count of Collision Pairs Before and After (Kepong RNCs) 180

• Comparison of plans: •

140 120

•

72

80 60

290 to 136

37

40

Macro – Indo or < 2km q

106

100

Macro – Macro < 5km q

20

6

12

0

0 k m < = x < 1 k m

35

40

(SC plan imp roved fr om 37 to 0 collisi on in th is category)

Current

New

5

10

15

20

25

30

12 0

Count of Collisions < 2km, Outdoor - Indoor

0

169

Count of Collision Pairs (Macro to Macro)

160

12

0 1 k m < = x < 2 k m

2 k m < = x < 3 k m

3 k m < = x < 4 k m

4 k m < = x < 5 k m


SbgHiTec h

• SC collision detection SbgHiTech RNCs were done by Optimizer. • Collisions were fixed by the NSN optimization teams using their in-house macro . • Total no of SC changes : 364

Interference Analysis

Interference Analysis • Interference analysis was done for both Kepong and SbgHiTech RNCs separately. • Optimizer was used to detect the overshooting cells and tilt recommendation was done manually upon having the site information from Celcom pertaining to these cells. • Each and every cell ,that was detected as potential overshooting cell, was checked individually to decide on the severity . • The most severe ones were considered for the tilt recommendation Overshoot ing cell count :

Tilt recommendation sample

Area

No o f o ver sh oo ti ng Cell s

Kepong

54

SbgHiTech

27

Interference Analysis Sample o vershoot ing Cells f rom K epong

Kepong

Interference Analysis Sample o vershoot ing Cells from SbgHiTech

SbgHiTec h

Interference Analysis Exception :

• Upon request from the NSN project team , the tilt recommendation for Kepong area was done by the Optimizer tools team with their limited know ledge about the geography of the target area.

• Celcom was not convinc ed with the recommendations made because of the same reason. • Hence it was taken over by the NSN local optimization team to revalidate the tilt recommendation before re-presenting it to Celcom again. • As agreed , the tilt recommendation for SbgHiTech area would be done by the NSN local optimization team.

Capacity Analysis

Capacity Analysis • Capacity Analysis was done for both Kepong and SbgHiTech RNCs separately. • Optimizer was used to detect the potential capacity issues in the network and generated the list of WBTS/WCELL affected . • Using the list, the local Optimization team came up with the solution plan. The analys is was done on 3 separate int erfaces

• Radio interface UL interference, DL Power limitation, Code occupancy,

• WBTS Channel Element Capacity Baseband Processing capacity

• Iub Transport Interface VCC Egress Utilization VCC Peak Capacity Utilization AAL2 Channel ID Utilization IP Iub Capacity

3G Optimizer for Celcom – Malaysia 38

© Nokia Siemens Networks

Optimizer: Radio Interface Rules •BTS Power Usage •Code Utilization •Users per Scheduler • DL/UL Load

From Optimizer, 2288 wcells identified for above criteria, in prior to list top worst cell, site list filtered by average code occupancy by > 95% on wcell level by hourly stats for 2 weeks ( 28/12/2011 – 11/1/2012)

Radio Interface Measurement Proactive Measurements

DL Power: Transm itt ed Total carri er power, Transmi tted non -HSDPA power UL interference: Received Total wideband power by the BTS Simultaneous number of HSDPA users Common channel Load (RACH, FACH, PCH) Channelisation Code util isation

Reactive Measurements

RRC/RAB, PS Call or HSPA setup failures Rejection rate in do wn lin k and u pli nk (SRB, AMR, CS data, PS data) Low HSDPA throughput Channelisation cod e blocking rate Radio Bearer dow ngrade or released due to congestion NRT DCH setup & reconfiguration rejection due to codes & power (DL & UL, RU10)

Analysi s cr it eri a * From optimizer, average code occupancy > 95% hourly wcell levell bi-weekly stats * RT, RRC , PS & HSDPA setup failure due to admission control > 0.3 % daily wbts level bi-weekly stats •Rejection rate PS DL/UL > 0.8 % daily wbts level bi-weekly stats •* HSDPA /HSUPA accessibility failure due to maximum user reached >0.5% daily wcell level bi-weekly stats •Failures due to admission control (AC) on the RRC, RAB and Channelisation Codes give a good representation of how severe the radio loading is on the network performance

Worst WBTS : Accessibil ity Failure due to AC(Admission control ) Sampl e sol uti on recommendation for RNC117 in Kepong WBTS

RNC

Worst Wbts List

STDNASIONAL

KPR17

High RRC Failure & Code ocupancy ( Optimizer)

ILSAS_BANGI

KPR17


C_PUTRAPERDANA

KPR17


SERDANGRAYA2

KPR17


C_BDRTASIKSLTN

KPR17


KJG_UTAMA

KPR17


Proposed Tuning

Parameter involved Maximum User reached Prxtarget,PrxMaxTargetBTS, PrxtargetPSmax With HSDPA enabled in the cell the code tree usage is already about 38%. The average code usage is higher than 70%. To set to Tune Prxtarget ( wcell 1,2,5,6) to improve 60%.Current value is 40%.If no improvement, the ps failure due to ac below 0.5 %. Tune DPCHOverHSPDSCHThreshold from 5 to HSPDSCH Code tuning lower No Maximum User HSDPA/HSUPA reached Prxtarget,PrxMaxTargetBTS, PrxtargetPSmax With HSDPA enabled in the cell the code tree usage is already about 38%. The average code usage is higher than 70%. To set to Tune Prxtarget ( wcell 3, 5,6) to improve the 60%.Current value is 40%.If no improvement, ps failure due to ac below 0.5 %. Tune DPCHOverHSPDSCHThreshold from 5 to HSPDSCH Code tuning lower Maximum User HSUPA reached Prxtarget,PrxMaxTargetBTS, PrxtargetPSmax With HSDPA enabled in the cell the code tree usage is already about 38%. The average code usage is higher than 70%. To set to Tune Prxtarget ( wcell 3,4, 5,6) to improve 60%.Current value is 40%.If no improvement, the ps failure due to ac below 0.5 %. Tune DPCHOverHSPDSCHThreshold from 5 to HSPDSCH Code tuning lower Maximum User HSUPA reached Prxtarget,PrxMaxTargetBTS, PrxtargetPSmax With HSDPA enabled in the cell the code tree usage is already about 38%. The average code usage is higher than 70%. The recommendation is to use 60%.Current value Tune Prxtarget ( wcell 5,6) to improve the ps is 40%.If no improvement, failure due to ac below 0.5 %. DPCHOverHSPDSCHTh resho No Maximum User HSDPA/HSUPA reached Prxtarget,PrxMaxTargetBTS, PrxtargetPSmax With HSDPA enabled in the cell the code tree usage is already about 38%. The average code usage is higher than 70%. The recommendation is to use 60%.Current value Tune Prxtarget ( wcell 7,8,9) to improve ac is 40%.If no improvement, failure 4% DPCHOverHSPDSCHTh resho Maximum User HSUPA reached Prxtarget,PrxMaxTargetBTS, PrxtargetPSmax With HSDPA enabled in the cell the code tree usage is already about 38%. The average code usage is higher than 70%. The recommendation is to use 60%.Current value Tune Prxtarget ( wcell 3,4,5,6) to improve is 40%.If no improvement, the ps failure due to ac below 0.5 %. DPCHOverHSPDSCHTh resho No Maximum User HSDPA/HSUPA reached

Action Item Prxtarget,PrxMaxTargetBTS, PrxtargetPSmax With HSDPA enabled in the cell the code tree usage is already about 38%. The average code usage is higher than 70%. To set to 60%.Current value is 40%.If no improvement, DPCHOverHSPDSCHThreshold from 5 to lower

Monitor after HSUPA parameter tuning

Monitor after HSUPA parameter tuning Prxtarget,PrxMaxTargetBTS, PrxtargetPSmax With HSDPA enabled in the cell the code tree usage is already about 38%. The average code usage is higher than 70%. The recommendation is to use 60%.Current value is 40%.If no improvement, DPCHOverHSPDSCHThresho

Monitor after HSUPA parameter tuning Prxtarget,PrxMaxTargetBTS, PrxtargetPSmax With HSDPA enabled in the cell the code tree usage is already about 38%. The average code usage is higher than 70%. The recommendation is to use 60%.Current value is 40%.If no improvement, DPCHOverHSPDSCHThresho

Worst WBTS : Acc essibility Failure due to AC(Admission control) Sampl e solutio n recom mendation for RNC118 in SbgHiTech

WBTS

RNC

Average Average Average Average Average of NRT Average Average Average of HSof HSof RRC of RT RAB RAB of PS Call of HSDPA of E-DCH DSCH stp DSCH stp setup FR, setup FR, setup FR, setup FR, stp FR,UL stp FR, FR, fails, DL AC AC AC AC DCH maxUsers maxUsers ovld

Maximum User reached

Action Item

SEK20_SHLM

118

0.095071 0.420459

0 0.286891 6.41E-05 0.008314

0

0 Maximum User HSUPA reached


KM14_FED_HWY

118

0.047716 0.492534

0 0.345788 0.000378

0

0

0 No Maximum User HSDPA/HSUPA reached

Tune Prxtarget

TMNSRIANDALAS2

118

0.043631 0.107418

0 1.282417

0

0


Tune Prxtarget

UITMKEJUTERA AN

118

0.042683 0.013984

0 0.892203 0.108914

0 Maximum User HSDPA reached


C_MASCOMPLEX

118

0.028452 0.036764

0 0.700651 0.271335 3.086079



C_TMSGRENGGAM

118

0.027362 0.002077

0 0.455551

0 0.00973 0.003179

0 Maximum User HSUPA/HSDPA reached


PLBH_BARAT2

118

0.024716

0 0.329108

0

SUBANGBESTARI

118

TMNBKTRIMAU

0 0.015815 0

0

0


Tune Prxtarget , SF128 tuning

0.022423 0.000278

0 0.404283 2.17E-05 0.023291

0



118

0.019769 0.000814

0 0.796917

0

0


Tune Prxtarget

C_TMSRIANDALAS

118

0.009288 0.007314

0 2.324622 0.000109 15.10847

0



TMN_BAYUPERDANA

118

0.000603 0.002636

0 1.738224

0



C_SEK13SA

118

0.003395 0.010894

0 1.395511 2.55E-05

0.21258

0



C_PANDAMARAN

118

0.006934 0.039867

0 1.258346

0 0.018525

0



PUNCAKPERDANA

118

0.005253 0.001552

0 0.298709 0.002265 0.316295

0.00035 0.002991 Maximum User HSUPA/HSDPA reached


UITM_TERATAI

118

0.005502 0.034069

0 0.646882 0.001645 0.001782 0.001321 0.002564 Maximum User HSUPA/HSDPA reached


RAMOSEK7PLUMBUM

118

0.012439 0.003016

0 0.316276

0.00109 0.046019

0



JLN_SGUDANG

118

0.004179 0.000557

0 0.309205

0 0.024504

0



SEKU5_2_SHLM

118

0.002333 0.005324

0

0.15

0 0.024282

0



C_HWSELATKLANG

118

0.012835 0.000313

0 0.341401

0 0.007254

0



JLNZAMRUDSEK7SA

118

0.013276 0.000423

0 0.602509

0 0.000761

0



PULAUINDAH

118

0 0.163776

0

0


Tune Prxtarget

0.00259

0

0

0

0

0 0.000718

0

Optimizer: WBTS Interface Rules Proactive Measurements Maximum Available/ used CE

•CE Usage •Accessibility failure due to WBTS

Aver age rat io of CE us age HSUPA CE usage Reactive Measurements RRC/RAB, PS Call or HSPA setup fail ures due to WBTS BTS_HSUPA_HW_Not_limited_DUR BTS_HSUPA_HW_Limited_DUR BTS_HSUPA_NO_HW_CAPA_DUR

WBTS Interface : Analysis flow

Analys is criter ia

* Maximum Used UL CE seems to be bottleneck : Filtered by < 100 CE left after maximum used UL CE over Maximum available CE * Maximum used ratio UL/ DL > 85% •Service Accessibility Failure > 0.2 % •HSDPA access failures due the BTS are mainly related to HSDPA UL

Worst TOP 10 WBTS : Acc essibility Failu re due to WBTS & High CE Usage Sampl e sol uti on recommendation for RNC119 in Kepong Average Max of Average Max of Max of of Avg Avg of Avg Avg Max of Max of Max of Max Ratio of Ratio of Ratio of Ratio of Max Max Max Used Used CE Used CE Used CE Used CE LCG Availabl Used CE Used CE CE's for DL in for DL for UL for UL Group e CE DL UL ratio BTS in BTS2 in BTS2 in BTS

RNC

WBTS

KPR19

C_CHERASJAYA1

1

468

317

407

86.97

40.98

58.12

48.50

70.51

Parameter tuning

KPR19

C_CHERASJAYA1

2

756

490

600

79.37

48.27

58.33

54.46

69.84

Parameter tuning

KPR19

C_PANDANJAYA

1

490

333

436

88.98

40.16

60.61

46.24

74.9

Parameter tuning

KPR19

C_PANDANJAYA

2

720

474

611

84.86

49.33

61.11

55.93

74.86

Parameter tuning

KPR19

C_TMNMELATI

1

504

308

371

73.61

31.49

49.80

36.51

58.93

Check planned CE and CE Upgrade

KPR19

2

504

440

504

100.00

63.29

79.17

71.55

92.86


KPR19

C_TMNMELATI C_TMWANGSAMAJ U

1

800

557

743

92.88

44.40

59.62

50.25

77.5


KPR19

C_WISMAKAUTHAR

1

360

326

360

100.00

56.98

81.48

66.31

96.67

Parameter tuning

KPR19

C_WISMAKAUTHAR

2

864

517

672

77.78

45.02

54.51

51.87

68.63

Parameter tuning

KPR19

JLN_REJANG_STPK

1

612

549

612

100.00

75.60

84.48

88.98

98.69

CE Upgrade

KPR19

PSTBDRMELAWATI

1

612

481

608

99.35

53.29

68.79

60.85

84.31

CE Upgrade

KPR19

PUTRASULAIMAN

1

360

265

315

87.50

41.93

57.22

46.53

66.11

Parameter tuning

KPR19

PUTRASULAIMAN

2

864

442

500

57.87

35.85

43.98

39.23

50.46

Parameter tuning

KPR19

C_PANDANINDAH

1

490

316

398

81.22

40.78

55.92

47.12

68.78

Parameter tuning

KPR19

C_PANDANINDAH

2

720

488

610

84.72

50.19

61.25

58.56

74.72

Parameter tuning

KPR19

C_WISMAKAUTHAR

1

360

326

360

100.00

56.98

81.48

66.31

96.67

Parameter tuning

KPR19

C_WISMAKAUTHAR

2

864

517

672

77.78

45.02

54.51

51.87

68.63

Parameter tuning

Action Item

Worst TOP 10 WBTS : Acc essibility Failu re due to WBTS & High CE Usage Sampl e solut ion recommendatio n for RNC121 in SbgHiTech

Averag Averag Max of Averag Free CE Max of e of Max of e of Max e of Max of over max Max Max Max Max Used Avg Avg availble Used Used Used Used CE's free free over max CE DL CE DL CE UL CE UL ratio CE's CE's2 used CE

No

WBTS

RNC

Max of Max Availabl e CE

1

TMN_SEROJA

SHR21

543

438

311

548

347

100.92 276.13

369

112

Monitor for CE upgarde

2

DESA_MAHKOTA

SHR21

544

426

302

534

328

98.16 292.46

374

10


3

KM270_TIROI

SHR21

544

438

319

528

355

97.06 267.63

380

16

CE Upgrade

4

LABOHANDAGANG

SHR21

1224

428

348

491

379

95.90 511.93

1025

733

NIL

5

KM290BKTUNGGUL

SHR21

580

447

353

528

389

95.22 235.97

330

52


6

ERL_SEPANG

SHR21

700

510

411

665

500

95.00 281.91

447

35

CE Upgrade

7

BDR_SEPANG

SHR21

512

389

268

478

287

93.36 290.90

366

34

CE Upgrade

8

BDRUNIVMANTIN

SHR21

864

433

295

535

338

92.50 339.13

608

329

Parameter tuning on LCG1

9

BDRSERIPUTRA

SHR21

720

530

362

657

426

91.25 257.10

392

63

Parameter tuning

10

SAP_BB_NILAI

SHR21

580

418

328

496

362

91.18 262.46

360

84


Action Item

Optimizer: Iub Capacity Rules

Proactive Measur ements Optimizer : Traffic load measurements rate per VCC (cells / measurement period) • User plane load • CNBAP load • DNBAP load • AAL2 signaling load CAC resource reservation (DL) to check average reservations levels • AAL2 path average reserved bandwidth • Available Iub transport capacity for HSDPA Number of AAL2 connections per VCC • Check that CIDs are not running out Shared HSDPA AAL2 allo cation u til ization [%] • Check if SharedHSDPAAllocation is under utilized

Reactive Measur ements

•Blocking •AAL2 Resource reservation success rate •AAL2 CaC rejection •Failures •Physical transmission failure •Alarms

Iub Interface Capacity Analysis Analysi s cr it eri a * From optimizer, VCC Average Egress utilization rule (Average traffic load) > 50% 183 wbts identified, * RT, RRC , PS & HSDPA setup failure due to IuB > 0.5 % daily wbts level bi-weekly stats • AAL2 CAC Resource reservation , AAL2 CAC rejection > 0.5 % Hourly bi-weekly stats • AAL2 connection > 240 CID > 5triggering hourly bi-weekly stats • VCC & VPI utilization > 90% hourly bi-weekly stats

Optimizer List: WBTS identified high VPI Utilization Sample solutio n recommendation for Kepong WBTS

RNC_IF_VPI

Max of Egress

Average of Egress

Max of Ingress

Average of Ingress

Average of Egress2

Max of Egress2 hourly

% Max VPI Usage

Transport Failure

Blocking ~ AAL2

AAL2UPPCR01E gr

# E1

TUNSRILANANGUKM

KPR1715017

13942

6316.50

3272.00

1183.57

35.77

78.94

85.00

Spike in 29/1/2012

minimal alcap sig blocking ~0.1%

16402

4

BKT_KOMANWEL

KPR1715225

22123

8719.33

4374.00

1565.18

24.48

62.10

77.37

Spike in 7/1/2012

no blocking

28595

7

Action Item To monitor closely To monitor closely

SG_LONG4

KPR191113

11655

5276.14

2731.00

1123.70

40.06

88.49

95.32

Spike in 9/1/2012

ALCAP blocking(2/1-5/1/12)

12227

3

E1 upgrade/ ip migration

C_TMNMELAWATI2

KPR1913621

15907

8893.76

3350.00

1430.27

50.36

90.07

96.98

Spike 4,9/1/12, 28/12/11

no blocking

16402

4


TMN_BKT_TERATAI

KPR1914915

11742

7546.23

3739.00

1466.40

57.29

89.15

96.03

Average 0.1% Daily

minimal alcap sig blocking below ~0.1%

12227

3

MRR2_KARAKITC

KPR201005

11807

7072.32

3855.00

1285.58

39.38

65.74

99.44

Average 0.03% Daily

High CAC rejection ( average 0.05%)

11874

3

TMBTARANG

KPR201015

11562

4906.39

1925.00

809.87

36.42

85.83

94.56

Spike 6-10/1/12

minimal alcap sig blocking ~0.1%

12227

3

Spike in 7/1/2012

ALCAP SIG blocking, CID reach maximum during peak hour

11874

3

8050

2

KGSGLANG

KPR201088

11912

7162.15

3851.00

1246.57

39.88

66.32

100.32

E1 upgrade/ ip migration To monitor closely.To tune ebs parameter to improve CAC rejection To monitor closely E1 upgrade/ ip migration To monitor closely

KM435RAWANG

KPR201123

5918

1235.95

1095.00

288.77

13.76

65.90

73.52

Spike in 5/1/2012

PERODUA_RAWANG

KPR201124

27116

14315.00

4927.00

2200.00

39.89

75.48

81.64

High Transport failure > 10%

No blocking No blocking AAL2 resource reservation

33213

8

ok

SERENDAHTOWN

KPR201132

25184

9175.95

3929.00

1541.87

25.54

70.11

76.03

Average 0.03% Daily

no blocking

33123

8

ok

C_AWANARESORT

KPR201151

15193

3654.67

2973.00

869.25

20.35

84.59

92.63

Spike in 9/1/2012


16402

4


C_GENTINGVIEW

KPR201153

9474

1837.45

1624.00

379.28

10.23

52.75

57.76

Spike in 9/1/2012

no blocking

16402

4

C_TMULUKALI

KPR201183

11816

9310.41

4294.00

1859.71

69.11

87.72

96.64

Spike in 9/1/2012

no blocking

12227

3

JLN_RAJACHULAN

KPR2410623

11419

2937.42

2746.00

774.46

22.30

86.70

93.39

Spike in 9/1/2012


12227

3

C_TMTTC

KPR2411817

16737

6935.70

3620.00

1373.98

31.31

75.56

81.32

Spike in 7/1/2012

no blocking

20582

5

ok To monitor closely.To tune ebs parameter to improve CAC rejection To monitor closely.To tune ebs parameter to improve CAC rejection To monitor closely.To tune ebs parameter to improve CAC rejection

JLNPRAMLEE

KPR241189

10806

3560.46

2934.00

978.02

9.99

30.33

32.62

Spike in 6 ,10/1/2012

no blocking, spike 2/6/1/12)

33123

8

ok

C_TMKLTOWER

KPR2412213

23388

10537.18

5356.00

2278.69

29.58

65.65

70.61

Spike in 6 -9,10/1/2012

No blocking

33123

8

ok

KL_PAVILION

KPR241227

9121

2566.37

2638.00

799.79

19.49

69.25

74.60

Spike in 9/1/2012

no blocking

12227

3

ok

Output: WBTS identified high VPI Utilization Sample sol ution recommendation for SbgHiTech Average of Ingress

Average of Average of Average of Egress Ingress2 Egress2

Max of Egress2

AAL2UPPC Performance R01Egr hourlyVPI utilization

RNC_IF_VPI

WBTS

SHR11_104_19

C_TMPJCENTRAL

2131.695522 8941.716

67.88946

91.26

12227

nok

E1 upgrade / ip migration

SHR18_116_19

C_SEKU8SA

1923.360119 8223.744

62.43842

90.68

12227

nok


SHR11_134_19

C_PUCHONGJAYA2

1273.271642

5257.17

60.56675

90.62

8050

nok


SHR11_139_21

NPE_PANTAI2

1938.313433

10869.5

61.5451

90.5

16402

nok


SHR18_123_5

JLNKEBUN

1897.050595 9423.967

71.55089

90.43

12227

nok


SHR18_103_19

C_SRIMUDA

67.95

90.4

8050

nok


1211.0625

5898.06

Action Item

SHR18_122_3

JLNKEBUN2

1737.720238 9421.628

71.53342

90.4

12227

nok


SHR18_106_5

KM12FED_HWY

921.0833333 4210.512

48.50792

90.38

8050

nok


SHR18_112_13

PULAUINDAH

1519.139881 6819.318

78.56378

90.38

8050

nok


SHR18_117_7

SEK30SHAHALAM

852.4047619 4861.119

56.00339

90.37

8050

nok


SHR11_152_9

FEDHWY_LDPITC

2575.910448 12767.44

72.29179

90.33

16402

nok


SHR18_111_3

C_SPG4PULAI

1664.940476 8188.122

62.16765

90.31

12227

nok


SHR18_103_21

C_KOTAKEMUNING

1357.369048 7265.307

55.1614

90.29

12227

nok


SHR11_112_3

SGPENCHALA2

1396.077612 7572.191

57.49161

90.23

12227

nok


SHR11_139_13

C_LDGKINRARA

1079.41194

4499.257

51.83484

90.18

8050

nok


SHR18_107_19 SHR11_116_7

PLBH_BARAT2 C_SGBULOHIP

1200.369048 5930.688 869.3522388 4998.307

68.32586 37.94913

90.18 90.15

8050 12227

nok ok

E1 upgrade / ip migration NIL

SHR11_138_9 SHR11_148_3

C_SS5PJ C_BT5PUCHONG

1450.310448 7684.155 1099.629851 6461.466

58.34191 49.05788

90.14 90.13

12227 12227

nok ok

E1 upgrade / ip migration NIL

SHR18_107_33

C_PANDAMARAN

1850.845238 7388.774

56.0989

90.12

12227

nok


SHR11_101_23

PELANGI SQUARE

1537.602985 7350.555

55.80863

90.11

12227

nok


SHR11_141_15

TMNKANAGAPURAM

1760.408955 7360.731

55.88546

90.06

12227

nok


SHR21_108_6

BDRBKTMAHKOTA

2905.342262 13511.96

75.22979

89.25

16402

nok


SHR21_105_16

TMN_PELANGI_SEM

2257.949405 9869.063

73.26149

88.51

12227

nok


SHR21_102_7 SHR21_101_14

SEMENYIH TLK_PGARANG

1903.434524 10061.78 2422.5625 18392.68

56.02042 51.20027

87.94 87.86

16402 33123

nok nok

E1 upgrade / ip migration IP migration

SHR21_111_6

TM_JENJAROM

1573.952381 7340.824

54.49351

87.8

12227

nok


SHR21_106_8 SHR21_104_3

TMN_GADONGJAYA BT8_SIJANGKANG

1314.10119 7288.271 948.6636905 6152.693

54.10351 45.67339

87.42 87.18

12227 12227

nok monitoring

E1 upgrade / ip migration Monitoring

SHR21_103_28

BANTING

1863.443452

56.53821

86.91

12227

nok


7616.28

OJT • Considering the amount of work involved in projects using Optimizer, it was suggested to have more manpower trained on this tool • Teton Escobar from Philippines was nominated for the OJT • Teton was involved in the Optimizer activities and had hands on training on the tool • The OJT spanned over for more than 3 weeks in Jan 2012.

Issues faced Connection Speed iss ue

• It was observed that the connection speed to Netact was very slow from Crest . • Other means were tried to connect remotely to Netact ,which didn’t help much. • Finally, it was arranged to continue with the work being based in Kepong itself. • All these delayed the project delivery timeline badly during the initial days of the project. ADJS KPI retrieval issue

• ADJS level KPIs were not available in the system because the license for M1028 counter was not available • The issue was resolved after installation of the license.

Conclusion and Way forward • The Antenna DB has to be updated on regular basis • While the KPI monitoring post nbr addition for Kepong showed major improvement, no remarkable improvement observed for SbgHiTech RNCs. The KPI monitoring for the same is going on.

• Neighbour optimization has to be done in regular intervals. • It is suggested to use Optimizer for SC allocation to get the optimum result instead of having grouping pattern as followed by Celcom.

• The Proactive approach to detect the capacity issues in Optimizer does not suit the current needs in Celcom. The reactive approach is preferred and recommended to be added in Optimizer.

3G OPtimizer 3.0 for Celcom

Recommend Documents