Radio Network Design and Optimization Department
Security Level: Confidential
CMPak Swap Project Cluster Acceptance Procedure and Criteria
2016-11-15
Radio Network Design and Optimization Department
Page1, Total9
Radio Network Design and Optimization Department
Security Level: Confidential
1
Introduction & Objective............................................................................................................3
2
Cluster Acceptance.......................................................................................................................3
3
Reporting and Submission for Acceptance...............................................................................
!
"#clusion Criteria........................................................................................................................$
2016-11-15
Radio Network Design and Optimization Department
Page2, Total9
Radio Network Design and Optimization Department
Security Level: Confidential
1 Introduction & Objective The objective of this document is to provide CMpak with the procedures and criteria for Acceptance of CMpak Swap Project. This would include comparison of cluster performance before and after the swap. It also addresses some exclusion criteria that would be applied in order to lter out the cells that are performin! poorl" due to issues out of the project scope.
2 Cluster cceptance PAC will be based on cluster level comparison of #SS statistical and $rive Test %PIs before and after swap. Acceptance items and criteria of OSS statistical KPIs:
The followin! table shows the acceptance items of #SS statistical %PIs for &MTS technolo!" as per 'PM contract.(su!!est dail" %PI acceptance due to )amadan * UMTS acceptance OSS KPIs for Swap
S N
KPI
Before Swap
After Swap
Before Swap
After Swap
RRC Setup Success Ratio (%) RAB Setup Success Ratio (%) CS Service Drop Ratio (%) PS Call Drop Ratio (%) Soft HO Success Ratio (%) CS W2G Inter RA! HO out Success Ratio (%)
1 2 3 ! %
UMTS monitor OSS KPIs for Swap S N 1 2 3 ! % $
KPI HSDPA "ser Avera#e !$rou#$put (&ps) CS traffic PS traffic R!WP "' C utiliation HSDPA "sers HS"PA "sers !CP utiliation
Table 1 UMTS OSS KPIs Acceptance items 2016-11-15
Radio Network Design and Optimization Department
Page3, Total9
Radio Network Design and Optimization Department
Security Level: Confidential
+or #SS %PI acceptance, the network should be fro-en for / da"s before swap.
+or #SS %PI comparison, the pre0Swap %PI value shall be the avera!e value of 12 da"s and the post0swap %PI should be the avera!e value of 3 da"s. A cluster shall be acceptable for CMpak when the followin! #SS %PIs are e4ual to or better than that before swap.
+or UMTS monitor OSS KPIs need to be as monitor %PI instead of acceptance %PI, if have lar!e 5uctuation we can clarif" it b" observe 6&7 tra8c to!ether
Acceptance criteria for T KPIs:
In !eneral, a cluster performance after Swap would be considered acceptable as lon! as there is no loss of covera!e when compared with pre swap drive test. +or ver" few points de!raded In post $.T, althou!h in total the $.T is better than before and these points not de!raded too much, it should not a9ect cluster acceptance
In order to compare the pre and post0drives, Cluster drive tests will be conducted with same route, with the same e4uipment and settin!s.
After cluster swap, we can or!ani-e post0$T at once.
As per CMPA% re4uired use +1:+3 !enerate new dual0carriers instead of old network fre4uenc" usin!, but $T test take free mode (don;t lock fre4uenc"*, so the pre and post $T covera!e will have sli!ht di9erentia, the phenomenon should be acceptance.
T!e ot!ers acceptance criteria statement
<# feedback within 1 da" and acceptance report (cluster* feedback within da"s.
It is no limitation for parameter chan!e in swap da" due to some ur!ent activities, after that we will make record and send them. =uawei would not re4uire an" <# to take such corrective measure.
Swap sites if belon! to indoor site which just cover indoor area suppl" fundamental service verication and I>S sites should eliminate from cluster ran!e due to implement and install hard can;t catch swap pro!ress
2016-11-15
Radio Network Design and Optimization Department
Page, Total9
Radio Network Design and Optimization Department
Security Level: Confidential
Some secretl" re!ion or speciall" re!ion which can;t accessed since emer!enc" like road blocked, protest activit", heav" rain etc. which !enerate pre and post covera!e have a di9erent. +or those case just clarif" it with photos, poster, event, news and shouldn;t a9ect swap cluster acceptance. After cluster acceptance will automaticall" be handed over to CMPak:'PM teams. +or user complain if happened after cluster swap and #SS %PI under monitor period, =uawei is char!e of complain handlin!. )+ acceptance shouldn;t relate to hardware part, the" can;t be conditional on to >SS or an" other acceptances. CMpak suppl" free SIM card and =i!h performance +TP. $urin! swap if individual sites remained in swap cluster due to proprietor blockin! or permission issue and !enerate cross sites between =uawei and ?T@, but over B sites had completed swap, we can implement Post0test. #nce the $T %PI and #SS %PI a9ected b" crossed sites we need to lter those cells to calculate %PI but if still low than pre0%PI the cluster should be acceptance accordin! clarication.
Test met!ods and test tools description for eac! ser"ice
Test Tools description
Tools escription T@MS 1D S2 6PS Sierra wireless 6 watcher Call test 'o. +TP Server
Comments &sed for $rive Test (6* Support 6 technolo!", CS service, E S2 6A)MI' &sed to test PS Service $estination no.( S2* =i!h performance +TP
&MTS $.T %PI and covera!e %PI as per 'PM re4uirement Cl#ster NO$
T KPI
Pre%test KPI
Post% test KPI
CS CSSR CS AMR Call Drop Rate CS Soft Handover Success Rate PS RAB Assignment Success 2016-11-15
Radio Network Design and Optimization Department
Page5, Total9
Radio Network Design and Optimization Department
Security Level: Confidential
rate PS Drop rate HSDPA Moving Throughput Average RSCP ( Idle Average !C"I# ( Dedicated
Items
U&'(U&)*+ock,- .#t don/t lock fre0#enc12 Common Settin!s
Mobile Placement Test t"pe
7aptop Foice
Mobile t"pe
Items Common Settin!s Mobile placement Test t"pe
ata Card )
7aptop =S$PA
Mobile t"pe
$uration
1/ sec
Idle period
1Bsec
Call number
+ile Si-e
2GD 6b
Idle period +TP Server
6PS 6PS
6PS 6PS
Software
Software
$rive test tool
$rive test tool
Post processin!
Post processin!
! Reporting and "ubmission #or cceptance A cluster acceptance procedure will be initiated as soon as the sites of that cluster have been swapped. The %PIs for acceptance will be monitored on hourl" basis and to verif" the basic functionalities and performance. A report on cluster level major %PIs will be sent to CMpak within a da". Afterwards, dail" level stats monitorin! and optimi-ation will be done to ensure that there is no de!radation in performance compared to pre swap %PIs. As soon as the %PIs are stable and all acceptance criteria have been satised, the report for PAC will be submitted to CMpak, the cluster will be considered provisionall" accepted b" CMpak if no issues are fed back from CMpak in workin! da"s. Please refer to the below attachment HCluster Acceptance )eport Template; for further details re!ardin! the cluster acceptance report format for each technolo!". Sin!le Site Ferication template is also attached below for reference.
2016-11-15
Radio Network Design and Optimization Department
Page6, Total9
Radio Network Design and Optimization Department
Security Level: Confidential
6 Swap Cluster Acceptance )eport.d
2016-11-15
+unctionalit" Test Template.xls
Radio Network Design and Optimization Department
Page!, Total9
Radio Network Design and Optimization Department
Security Level: Confidential
$ %clusion Criteria An" de!radation in performance due to below mentioned scenarios shall be excluded from %PI calculation for #SS and $T %PIs
Since we are usin! +1:+3, the throu!hput results would be di9erent when compared to cells usin! +1:+E. @xclusion needs to be !iven in such cases.
The hardware issues, software or parameter con!uration of other vender, core network issues (especiall" #S, C6I, pa!in! issue * or other issues related to end to end connectivit"
The period where major network activit":retunin!:chan!e:incident is performed should be excluded from the %PI calculation with mutual a!reement.
Chan!e in tra8c due to user behavior, special a!!ressive promotional packa!es
=i!h tra8c @vents adversel" a9ectin! performance.
7=) network almost dual0carries covera!e but exist the two carries discontinues covera!e area. #nce user lon!0term random occup" discontinues fre4uenc" which will cause di9erence in $T covera!e and %PI between Pre0swap and Post0 swap. So once the phenomenon is appeared we just make a clarication
The I&) strate!" will be implemented durin! 6 network swap. Since I&) parameter wron! con!uration, no I&) deplo"ment or I&) parameters not con!ured even I&) compatibilit" issue between =uawei and ?T@ cause the mar!inal de!radation on a specic %PI such as S=# S), tra8c, C$) which should be excludedJ ?T@ should be char!e of correspondin! I&) strate!" and parameter desi!n even implement verication test for I&) compatibilit" at ?T@ side, if due to ?T@ side I&) con!uration issue cause %PI de!radation and complaint should be excludedJ
@xternal interference or >urst lar!er re!ion interference should be as clarif" condition
?T@ have a lar!e chan!ed such )'C level parameters modied, reframin!, 6 power adjust, the second carries expanded, up!rade action durin! swap or network free-e cause the %PI
2016-11-15
Radio Network Design and Optimization Department
Page", Total9
Radio Network Design and Optimization Department
Security Level: Confidential
mappin! result mismatchin! make %PI !o down should be as clarif" condition
if %PI have a de!radation due to transmission related issues includin! but not limited T) alarms, T) #S issue or bandwidth limitations should be as clarif" condition
$urin! swap the pa!in! rate ma"be have a sli!ht 5uctuation due to few sites swapped but received much pa!in! messa!e, this case shouldn;t be critici-ed
Some ?T@ cell con!ured CPIC= power over <, we need to fellow accordin!l".
2016-11-15
Radio Network Design and Optimization Department
Page9, Total9
