India VDF 4G MS Team Optimization Process v1.3
India VDF 4G MS Team Optimization Process v1.3
India VDF 4G MS Team Optimization Process v1.3
Product version
Total 19 pages
V1.0
Reviewed by Date
Reviewed by Date
Approved by Date
The L.RRC.SetupFail.Rej counter is incremented by 1 each time the eNodeB sends an RRC
Connection Reject message to the UE after receiving an RRC Connection Request message
from the UE. If the reject reason is due to resource allocation failure then the
b. Check <L.Traffic.User.Max> if reach the product caps or the license limited <RRC
Connected User(per RRC Connected User)> ---normally it’s 400 per site.
DSP License:
c. Check CPU load < VS.Board.CPUload.Mean>, normally it should be less than 70%.
e. Check configuration parameter : Use FMA tool to compare with normal site.
f. Get board log and use FMA to check if any internal fault in the cfltlog.log. like <low-
layer link of S1 interface fault>, need push BSS & transmission team to check.
setup, or delay to reply/UE didn’t receive RRC setup MSG/eNodeB didn’t receive RRC setup
a. Check any alarm like ALM-26529 RF Unit VSWR Threshold Crossed/26532 RF Unit
Hardware Fault/26200 Board Hardware Fault. Other alarms which time is mapping with
LST CLKMODE : to check GPS clock mode, if free will cause high interference.
WaitRrcConnSetupCmpTimer (Eran 6). Other parameter setting Use FMA tool to compare
e. Check CPU load < VS.Board.CPUload.Mean>, normally it should be less than 70%.
f. Put the UU trace to check signal flow to find which MSG is missing and calculate each
MSG time.
g. Coverage issue: Check from CHR log, any UL weak coverage and the TA distributing (TA
counter will also be available in ERAN6.1), and try to do the physical optimization or Power
optimization;
RAB.FailEst.SecurModeFail.
2.1 L.E-RAB.FailEst.MME
b. Push Ericsson team help to check from EPC side (AMBR/ARP& etc.);
2.2 L.E-RAB.FailEst.TNL
2.3 L.E-RAB.FailEst.NoRadioRes
DSP LICENSE:;
c. Check board CPU load < VS.Board.CPUload.Mean>, normally it should be less than 70%
2.4 L.E-RAB.FailEst.SecurModeFail
ThirdCipherAlgo=NULL;
ThirdIntegrityAlgo=NULL;
2.5 L.E-RAB.FailEst.NoReply
c. Top UE issue ;
4. PS Drop
TOP Cell condition: continuous 2 days NBH >1% & Failure number ≥ 5 or 1 day BBH PS drop
number ≥ 20
3.1 RF Failure
L.E-RAB.AbnormRel.Radio
ENodeBMaxRetxThreshold=Maxretx_Threshold_t32
For other parameter check, Pls use FMA tool to compare with normal sites;
e. Check ANR switch enable or not and check any missing neighbor;
f. Weak coverage: need to check from CHR log, if have many ul weak coverage & UL signal
instant deterioration . if these two reason percent is high, need to improve the coverage;
5. HOSR
TOP Cell condition: continuous 2 days NBH <99% & Failure number ≥ 10 or 1 day BBH PS drop
number ≥ 20
4.1 HO summary
In the current Huawei LTE system, the intra-frequency handover algorithm is triggered by
event A3 and the event reporting mode is event-triggered periodic reporting. When the
quality of a neighboring cell is higher than that of the serving cell by a preset value, event
A3 is triggered. According to 3GPP TS36.331, the conditions for event A3 are as follows:
Ocn is the cell-specific offset of the neighboring cell and is determined by the
CellIndividualOffset parameter. When the value is not zero, this parameter is
delivered in the measurement control message; when the value is zero, this
parameter is not delivered. The parameter value determines the occasion
when an intra-frequency handover is triggered.
Ofs is the frequency-specific offset (QoffsetFreq) of the serving cell and the
default value is 0. It is not considered in the evaluation for intra-frequency
handovers.
Ocs is the cell-specific offset (CellSpecificOffset) of the serving cell; the default
value is zero.
4.3 Ho Failure
executions attempt, and there is no related failure counters. So we use below formula to
(L.HHO.IntereNB.IntraFreq.ExecAttOut) - (L.HHO.IntraeNB.IntraFreq.ExecSuccOut) -
(L.HHO.IntereNB.IntraFreq.ExecSuccOut)
b. Check any related alarm and clear the alarm, especially PCI conflict alarm;
c. Check ANR switch enable or not and check any missing neighbor;
e. Check X2 link, if not configuration, then request BSS team for help; and then do the ping
test to check if the link is ok or not, if have issue ,take transmission team for help.
h. Weak coverage: need to check from CHR log, if have many ul weak coverage & UL signal
instant deterioration . if these two reason percent is high, need to improve the coverage;
6. Throughput
If we face a throughput issue, we first need to very clear about the issue description:
What: what kind of issue ? no speed or low speed? For special Server or all the server?
After collect above information, it will be easy to begin analysis throughput issue:
a. If issue only happen from a special time, then pay more attention to check any
b. If issue is only happen on a special user, then pay more attention to the user
SIM/device/laptop issue. We can switch SIM/device to verify. For single SIM issue, also
c. If issue happen on special site, then first need to check the configuration.
LST IPPATH & LST IPRT to check IP configuration (need to configure all the12 UGW
IPs)
LST VLANCLASS to check VLAN setting , VLAN ID need to map with transmission
MIMO(per Cell)(TDD)
e. Check CPU load < VS.Board.CPUload.Mean>, normally it should be less than 70%. And
f. Check RF condition:
If RSRP is good, but SINR is poor, need to check the PCI planning, and the neighbor RSRP,
if both 2 cells RSRP’s gap is below 3dB, then try to do optimization and give the a main
service cell;
If Both RSRP & SINR is good, then need to check the RB count, if RB utilization is high
(>70%), and average user number is high(>20), but throughput is low, then can consider
g. Do the UDP/Ping test, if UDP is also low or have packet loss or big delay, then need to
L.RRCRedirection.E2W - L.CSFB.E2W
a. Check any major alarm and any nearby sites down cause the coverage poor;
8. CSFB
As current setting, our CS call is fall back to 3G by redirection;
a. Make sure the SIM card enable CS call (confirm with core network side)
b. Make sure TAC/LAC mapping & DNS etc. are definition well in MME side;
9. Neighbor optimization
As our current setting, we enable the ANR automatically adding neighbor.
So for neighbor optimization, we pay more attention to Remove the useless neighbor:
Data source: Neighbor relation from CME or OMSTAR; 1 Week Neighbor HO from PRS;
Optimization principle :
b. Between the Neighbors, there are no Ho attempt in total 7*24 hrs & double-direction
d. Above principle, if the remain neighbor is still more than 50, then put the distance to
over 500M;
e. When make script , also make the double-direction Neighbor & also remove the
M2000.
a. Check the conflict neighbor distance & azimuth, if distance is more than 3KM/over 4
layer sites, and also coverage different area, then we can remove the neighbor;
b. If distance is near or coverage towards same area, then need to re-plan the PCI;
c. If after remove neighbor, it’s continuous repeat, then need to check the coverage and
KPI
Object hold
(Red)
License
L.Traffic.eNodeB.User.Max
usage of eNode >=80
/Maximum number of UEs add license
activated B %
specified by the license
Usage
=L.ChMeas.PRB.DL.Used.Avg/100
MOD
PRACH
CELLALGOSWITCH:
Resource (L.RA.GrpA.Att + >=75
Cell LocalCellId=x,
Usage(co L.RA.GrpB.Att)/3600/100 %
RachAlgoSwitch=Bac
mpete)
kOffSwitch-1
PRACH MOD
Resource CELLALGOSWITCH:
>=75
Usage(no L.RA.Dedicate.Att/3600/100 Cell LocalCellId=x,
%
n- RachAlgoSwitch=Ma
compete) ksIdxSwitch-1
1、MOD
CELLALGOSWITCH:
LocalCellId=xxx,
PucchAlgoSwitch=Pu
(L.ChMeas.CCE.CommUsed +
PDCCH cchSwitch-1;
L.ChMeas.CCE.ULUsed + >=80
Resource Cell 2、Physical
L.ChMeas.CCE.DLUsed) %
Usage optimization
/3600/1000/84
3、Divide the
coverage of the
cells
L.Thrp.bits.DL/ L.Thrp.Time.DL
/1000;
1.Check alarm or
RRC
L.RRC.SetupFail.ResFail / board fault;
Congestio Cell
L.RRC.ConnReq.Att 2.Physical
n Ratio
optimization;
1.Check alarm or
E-RAB
L.E-RAB.FailEst.NoRadioRes / L.E- board fault;
Congestio Cell
RAB.AttEst 2.Physical
n Ratio
optimization;
13. Alarm
Important alarm related to Access/HOSR/PS Drop/Throughput:
Alarm.xlsx