Sunday, 1 November 2015

LTE FDD Network Planning & Optimization System-Demo Cluster Area Optimization Report

1                 Overview

 

1.1              Test Purposes

 
The major purposes of this evaluation are to learn about the present coverage quality and running performance of the network in demo areas. On the basis of the network’s actual conditions, manage to solve the existing network problems and present suggestions for performance optimization especially the coverage and DL throughput performance.

1.2              Test Area

 
Demo Cluster include 4 sites , it is an urban area.


 

2                 Test Result Comparison

 

2.1              Test result overview

 
  Table 21  test KPI before optimization overview
 
Table 22  test result after optimization overview


 

2.2              RSRP

 

RSRP DT result before optimization
Figure 21  RSRP DT result at 6-19

  RSRP DT result after optimization
Figure 22  RSRP DT result at 6-22

 
Figure 23  RSRP Comparison
 
The RSRP comparison shows that the percent of RSRP poorer than -95 dBm has decrease from 31.79% to 26.20%, the coverage become better.
 

2.3              SINR

 
SINR DT result before optimization

Figure 24  SINR DT result at 6-19
 
 SINR DT result after optimization
 Figure 25  SINR DT result at 6-22

 
Figure 26  SINR Comparison

 
The SINR comparison shows that the percent of SINR poorer than 5 dB has decrease from 13.32% to 9.84%, the SINR become better.
 

2.4              MAC DL Throughput

 
MAC DL Throughput DT result before optimization
Figure 27  MAC DL Throughput at 6-19

 
MAC DL Throughput DT result after optimization
Figure 28  MAC DL Throughput at 6-22

 
Figure 29 MAC DL Throughput Comparison

 
The MAC DL Throughput comparison shows that the percent of MAC DL Throughput less than 20Mbps has decrease from 19.85% to 6.69%, the DL throughput become better.
 

3                 Optimization Case

 

3.1              Random Access Failure Case Analysis

 
Analyzing the DT log at 6-19, we found that there are random access failure cases.
Figure 31  Random access failure case
 
Checking the PRACH parameters and found that the parameter setting are not reasonable: neaby  cells shoud not have the same prachConfigIndex&rootSequenceIndex at the same time, otherwise it will cause cell confusion and result in random access failure.
 
ENBFunctionFDD
EUtranCellFDD
prachConfigIndex
rootSequenceIndex
5004
1
5
831
5004
2
5
801
5004
3
5
831
Optimization suggestion: re-plan prach parameters for relate cells.
 
Table 31  parameters optimization1

site ID
cell ID
parameter name
old value
new value
5004
1
prachConfigIndex
5
9
5004
2
prachConfigIndex
5
10
5004
3
prachConfigIndex
5
11
5004
1
rootSequenceIndex
24
618
5151
1
rootSequenceIndex
125
600
5151
2
rootSequenceIndex
133
606
5151
3
rootSequenceIndex
141
612

3.2             Call Drop Case Analysis

 
Analyzing the DT log at 6-19, most the call drop are caused by handover failure. Detail analysis and optimization suggestions please refer to section 3.3

 

3.3              Handover Failure Case Analysis

 

3.3.1                Handover failure caused by bad SINR

 

Here is one of the handover failure case analysis: UE at soruce PCI 43 found PCI 40 stronge enough and send Measurement Report, but UE did not receive handover command(RRC Connection Reconfiguration) and then call drop.
Figure 32  Handover failure case

 
The UE send the Measurement Report when serving cell SINR become very bad(-5.6 db).

 
UE campus on a new PCI 23 after call drop. From these procedures we can consider this handover and call drop are cause by bad SINR.

 
Optimization suggestion: optimization for SINR.
 

3.3.2                Handover failure caused by PCI 10 overshooting

 

UE catch a PCI 10 where is rather far away. PCI 10 is from site SoraviaRoof_SKO5556

 
UE failure handover to PCI 24 because PCI 10 overshooting and no neighbor relationship for this two PCI.

 
Another case is similar caused by PCI 10 overshooting:
UE catch signal from PCI 10 far away.


 
Optimization suggestion:
Add 2°downtilt for SoraviaRoof_SKO5556_2.
 

3.3.3                Handover failure caused by PRACH parameter 

 

From section 3.1, we know that 5004_1 and 5004_2 have the same PRACH setting, this can also affect handover.
The KPI from EMS show that handover success rate is not good for cell 5004_1 and 5004_2

 
Some handover also found related with these two sector during DT.

 
Optimization suggestion:
Adjust PARACH parameter as section 3.1
 

3.4             Coverage Case Analysis

 

3.4.1                 Bolnica_SKP5004_3 adjust tilt from 0°to 2°

 
Downtilt 0°may eaily caused overshooting to other cell, and according to DT result, area close to Bolnica_SKP5004_3 is not good enough, adjust tilt from 0°to 2°can enforce the coverage.
Figure 33  coverage analysis case2

 

3.4.2                 Divizija_SKO5073_3 adjust tilt from 0°to 3°

 
According to DT result, Divizija_SKO5073_3 is overshooting to site Bolnica.
Figure 34  coverage analysis case3

 

3.5              RS Power Optimization

 
After tilt change, increase the cell RS power can enforce the coverage.
Table 32  parameters optimization2
 


site ID
cell name
parameter name
old value
new value
5004
LTE1800_Bolnica_1
cellReferenceSignalPower
10.6
13.6
5004
LTE1800_Bolnica_2
cellReferenceSignalPower
10.6
13.6
5004
LTE1800_Bolnica_3
cellReferenceSignalPower
10.6
13.6
5073
LTE1800_Divizija_1
cellReferenceSignalPower
10.6
13.6
5073
LTE1800_Divizija_2
cellReferenceSignalPower
10.6
13.6
5073
LTE1800_Divizija_3
cellReferenceSignalPower
10.6
13.6
5169
LTE1800_MVRHQ_3
cellReferenceSignalPower
10.6
13.6

3.6              Handover Optimization

 
Comparing the RSRP and SINR, most the road RSRP is better than -95 dBm, UE can easily catch signal from neaby site. Optimize the parameter timeToTrigger from 7(256ms) to 6(160ms) can make the UE fast handover to a stronger neighbor, thus can let the UE stay at the strongest RSRP and SINR cell.
Optimization Suggestion:

site ID
parameter name
old value
new value
comment
5004
timeToTrigger
7
6
measCfgIdx=50
5073
timeToTrigger
7
6
measCfgIdx=50
5151
timeToTrigger
7
6
measCfgIdx=50
5169
timeToTrigger
7
6
measCfgIdx=50
 

Budi Prasetyo

About Budi Prasetyo

All About LTE

Subscribe to this Blog via Email :