#GOBEYOND

CUSTOMER
EXPECTATION

Adaptive Closed Loop MIMO (TM4)

Expand Trial

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
4G Strategic : Adaptive Closed Loop MIMO Expand Trial Proposal CUSTOMER
EXPECTATION

Fact Finding : Proposal

• 98.79% of the cells are configured with Adaptive TM3, even in low mobility area. • We proposed to implemented adaptive closed mimo with filtering some criteria below :
• We proposed to implemented on POI area which will be testing by Nielsen ( also exclude
MIMO Parameter MIMO setting Count of cells with criteria below ) :
InitialMimoType Adaptive 11658
No adaptive 11517 Number Filter
Mimo adaptive switch
Open loop adaptive 141 1 MIMO
2 Exclude VIP and VIP Route
3 RedCell by CQI
Background / Objectives: 4 Exclude Train Route
• By using TM4 it can increase the UE throughput by improving the precoding matrix 5 Inner Area
selection based on the UE feedback, but by sending the PMI uplink IBLER can be
CR Distribution
increased.
• Adaptive configuration of closed-loop transmission modes can be used to increase Row Labels Count of cells Batch Overview percentage cells jabo
performance if most UEs are moving at low speed, PMIs reported by UEs are reliable, and JAKARTA BARAT 184
demodulation performance is satisfactory in closed-loop transmission mode. JAKARTA PUSAT 155
JAKARTA SELATAN 222 1 ( 27 February 2017 ) 7.80
• When adaptive configuration of closed-loop transmission modes is enabled and some
JAKARTA TIMUR 178
UEs are discovered as incompatible UEs, the eNodeB can adaptively configure an open- JAKARTA UTARA 171
loop transmission mode for these UEs KOTA BEKASI 224
KOTA BOGOR 33
KOTA DEPOK 78
2 ( 1 March 2017 ) 11.41
KOTA TANGERANG 228
KOTA TANGERANG SELATAN 110
POI area 657
Grand Total 2240   19.21

Total around 19.21% proposed to be implemented using adaptive closed loop mimo in
Jabodetabek area

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
4G Strategic Expand Trial Result : Closed Loop Adaptive MIMO batch#1
CUSTOMER
EXPECTATION

Fact Finding : Parameter Setting :

• 98.79% of the cells are configured with Adaptive TM3, even in low mobility area MO Parameter Value
CELLMIMOPARACFG MimoAdaptiveSwitch CL_ADAPTIVE
Positive Results :
 Cell DL Avg and User DL Avg Throughput improve 8.83% and 6.67%
 CQI >7 Distribution improve 4.82%
 Other major KPIs are maintain

Background / Objectives:
• By using TM4 it can increase the UE throughput by improving the precoding matrix
selection based on the UE feedback, but by sending the PMI uplink IBLER can be
increased.
• Adaptive configuration of closed-loop transmission modes can be used to increase
performance if most UEs are moving at low speed, PMIs reported by UEs are reliable, and Negative Results :
demodulation performance is satisfactory in closed-loop transmission mode.  DL and UL IBLER degraded, especially on UL IBLER ( degraded by -2.14%), but not
• When adaptive configuration of closed-loop transmission modes is enabled and some impacted to UL throughput
UEs are discovered as incompatible UEs, the eNodeB can adaptively configure an open-
loop transmission mode for these UEs
Pros :
• Better channel estimation with closed loop MIMO
• Compare with Adaptive closed loop give better User Downlink throughput and UL IBLER
Cons :
• RI reporting may be reduced, Higher UL IBLER due to PMI reporting
Pre-requisite Execution Criteria :
• Expand trial at 910 cells on inner area on batch#1

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
4G Strategic Expand Trial Result : Closed Loop Adaptive MIMO batch#2
CUSTOMER
EXPECTATION

Fact Finding : Parameter Setting :

• 98.79% of the cells are configured with Adaptive TM3, even in low mobility area MO Parameter Value
CELLMIMOPARACFG MimoAdaptiveSwitch CL_ADAPTIVE
Positive Results :
 Cell DL Avg and User DL Avg Throughput improve 7.95% and 8.47%
 CQI >7 Distribution improve 5.58%
 Other major KPIs are maintain

Background / Objectives:
• By using TM4 it can increase the UE throughput by improving the precoding matrix
selection based on the UE feedback, but by sending the PMI uplink IBLER can be
increased.
• Adaptive configuration of closed-loop transmission modes can be used to increase
performance if most UEs are moving at low speed, PMIs reported by UEs are reliable, and Negative Results :
demodulation performance is satisfactory in closed-loop transmission mode.  DL and UL IBLER degraded, especially on UL IBLER ( degraded by -1.31%), but not
• When adaptive configuration of closed-loop transmission modes is enabled and some impacted to UL throughput
UEs are discovered as incompatible UEs, the eNodeB can adaptively configure an open-
loop transmission mode for these UEs
Pros :
• Better channel estimation with closed loop MIMO
• Compare with Adaptive closed loop give better User Downlink throughput and UL IBLER
Cons :
• RI reporting may be reduced, Higher UL IBLER due to PMI reporting
Pre-requisite Execution Criteria :
• Expand trial at 1306 cells on inner area on batch#2

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
4G Strategic Expand Trial Result : Closed Loop Adaptive MIMO – Batch#2 POI
CUSTOMER
EXPECTATION

Positive Results :
 Cell DL Avg and User DL Avg Throughput improve 9.43% and 30.07%
 CQI >7 Distribution improve 5.10% and DL PDSCH MCS improve 10.55%
 Other major KPIs are maintain

Negative Results :
 DL and UL IBLER degraded, especially on UL IBLER ( degraded by -1.63%), but not impacted to UL throughput

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
CUSTOMER
EXPECTATION

Adaptive Closed Loop MIMO (TM4)

Expand Trial

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
CUSTOMER
EXPECTATION

Outline
Summary & Suggestion

Improvement Result

Summary KPI Comparison

KPI Comparison

Additional KPI

Parameter/Feature Verification

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
Summary KPI Comparison Adaptive Closed Loop MIMO Batch #1 CUSTOMER
EXPECTATION

Classes KPI Before ( 21-22 February ) After (28 Feb - 1 March) Delta (%) Formula Remark
Rach Cont. Resol SR (%) 78.21 78.86 0.65 After-Before improved
RRC setup SR (%) 99.96 99.96 0.00 After-Before maintain
Accessibility
E-RAB setup SR (%) 99.91 99.92 0.01 After-Before maintain
Call Setup Success Rate (%) 99.67 99.65 -0.02 After-Before maintain
Retainability Service Drop Rate (%) 0.13 0.11 -0.01 After-Before maintain
Mobility Intra-Frequency Handover Out Success Rate (%) 99.74 99.81 0.06 After-Before maintain
CSFB CSFB Preparation Success Rate (%) 99.77 99.85 0.08 After-Before maintain
Cell Downlink Average Throughput 13.78 15.00 8.83 (After-Before)/Before*100 improved
Cell Uplink Average Throughput 1.78 1.90 6.67 (After-Before)/Before*100 improved
Service Integrity
User Downlink Average Throughput 7.98 8.98 12.53 (After-Before)/Before*100 improved
User Uplink Average Throughput 1.57 1.67 6.44 (After-Before)/Before*100 improved
DL Resource Block Utilizing Rate (%) 26.59 25.07 -1.52 After-Before maintain
Utilization
UL Resource Block Utilizing Rate (%) 15.53 14.58 -0.95 After-Before maintain
Average User Number 31995.94 32463.92 1.46 (After-Before)/Before*100 maintain
User
Maximum User Number 21344.42 21805.13 2.16 (After-Before)/Before*100 maintain
Downlink Traffic Volume 1402803.46 1483835.68 5.78 (After-Before)/Before*100 improved
Traffic
Uplink Traffic Volume 141118.64 144491.95 2.39 (After-Before)/Before*100 maintain
Availability Radio Network Availability Rate (%) 68.61 81.85 19.31 After-Before improved
Interference Average UL interference -106.48 -106.33 -0.15 (After-Before)/Before*100 maintain
CQI>7 (%) 80.75 85.57 4.82 After-Before improved
DL MCS 11.88 13.08 10.12 (After-Before)/Before*100 improved
Quality
UL MCS 15.16 15.24 0.54 (After-Before)/Before*100 maintain
MIMO Rank2 Ratio (%) 34.63 26.99 -7.65 After-Before degraded
DL IBLER (%) 11.33 11.98 -0.64 After-Before degraded
IBLER
UL IBLER (%) 14.34 16.48 -2.14 After-Before degraded

KPIs Maintain Improved Degrade Significant improvement:

Retainability [-0.03%,0.03% >0.03% <-0.03%  DL User Throughput
Accessibility & Availability [-0.07%,0.07%] >0.07% <-0.07%  DL Cell Throughput
Remaining KPIs [-3%,3%] >3% <-3%  CQI and DL PDSCH MCS average

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
KPI Comparison:: Traffic Volume and User ( Adaptive Closed Loop MIMO ) CUSTOMER
EXPECTATION

• DL and UL Traffic Volume having similar trend

• Average User Number also having similar trend.

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
KPI Comparison:: Downlink and Uplink Throughput ( Adaptive Closed Loop MIMO ) CUSTOMER
EXPECTATION

• Cell and User DL Throughput significantly increased around 8.83% due to higher CQI and MCS as per expected
• UL Cell and User Throughput showing stability
NETWORK QUALITY MANAGEMENT | 2017
 #GOBEYOND
KPI Comparison:: Accessibility and Retainability ( Adaptive Closed Loop MIMO ) CUSTOMER
EXPECTATION

• RACH Contention Resolution showing stability

• RRC, ERAB Setup Success & Service Drop Rate showing stability.

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
KPI Comparison:: Mobility and CSFB ( Adaptive Closed Loop MIMO – batch1 ) CUSTOMER
EXPECTATION

• IFHO and CSFB showing improve but not related to this activity

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
KPI Comparison:: PRB Utilization, CQI & Rank2 Rate ( Adaptive Closed Loop MIMO- batch1EXPECTATION
) CUSTOMER

• DL/UL PRB Utilization showing reduced but not related to this activity
• CQI >7 Distribution improve around 4.82% as expected
• MIMO
NETWORK rank-2
QUALITY are sdecreased
MANAGEMENT around 7% due to PMI reporting
| 2017
 #GOBEYOND
KPI Comparison:: IBLER, MCS and UL Interference ( Adaptive Closed Loop MIMO- Batch1 ) CUSTOMER
EXPECTATION

• DL MCS increased as expected around 10.12%

• DL and UL IBLER are decreased -0.64% and -2.14%
• UL interference show stable trend
NETWORK QUALITY MANAGEMENT | 2017
 #GOBEYOND
Additional KPI:: Related Feature/Parameter ( Fixed Closed Loop MIMO – Batch1 ) CUSTOMER
EXPECTATION

• Transmission Mode distribution switch between TM4 and TM2

• Rank distribution are changing from Open loop rank1 and rank2 into Closed loop rank1 and rank2

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
Summary KPI Comparison Adaptive Closed Loop MIMO Batch #2 CUSTOMER
EXPECTATION

Classes KPI Before (22-23 March ) After (1- 2 March) Delta (%) Formula Remark
Rach Cont. Resol SR (%) 83.40 84.39 1.00 After-Before improved
RRC setup SR (%) 99.98 99.98 0.00 After-Before maintain
Accessibility
E-RAB setup SR (%) 99.94 99.95 0.01 After-Before maintain
Call Setup Success Rate (%) 99.91 99.86 -0.05 After-Before maintain
Retainability Service Drop Rate (%) 0.08 0.08 -0.01 After-Before maintain
Mobility Intra-Frequency Handover Out Success Rate (%) 99.81 99.87 0.06 After-Before maintain
CSFB CSFB Preparation Success Rate (%) 99.83 99.91 0.07 After-Before maintain
Cell Downlink Average Throughput 14.56 15.72 7.95 (After-Before)/Before*100 improved
Cell Uplink Average Throughput 1.69 1.82 8.00 (After-Before)/Before*100 improved
Service Integrity
User Downlink Average Throughput 9.00 9.77 8.47 (After-Before)/Before*100 improved
User Uplink Average Throughput 1.57 1.71 8.56 (After-Before)/Before*100 improved
DL Resource Block Utilizing Rate (%) 27.92 26.66 -1.26 After-Before maintain
Utilization
UL Resource Block Utilizing Rate (%) 15.72 14.97 -0.75 After-Before maintain
Average User Number 24430.13 24175.16 -1.04 (After-Before)/Before*100 maintain
User
Maximum User Number 16379.30 16251.65 -0.78 (After-Before)/Before*100 maintain
Downlink Traffic Volume 1250400.89 1310838.21 4.83 (After-Before)/Before*100 improved
Traffic
Uplink Traffic Volume 105617.95 107568.03 1.85 (After-Before)/Before*100 maintain
Availability Radio Network Availability Rate (%) 84.65 99.64 17.71 After-Before improved
Interference Average UL interference -110.25 -109.80 -0.41 (After-Before)/Before*100 maintain
CQI>7 (%) 83.42 89.00 5.58 After-Before improved
DL MCS 12.45 13.84 11.21 (After-Before)/Before*100 improved
Quality
UL MCS 15.70 15.88 1.11 (After-Before)/Before*100 maintain
MIMO Rank2 Ratio (%) 38.31 30.24 -8.07 After-Before degraded
DL IBLER (%) 11.12 11.74 -0.62 After-Before degraded
IBLER
UL IBLER (%) 12.87 14.18 -1.31 After-Before degraded

KPIs Maintain Improved Degrade Significant improvement:

Retainability [-0.03%,0.03% >0.03% <-0.03%  DL User Throughput
Accessibility & Availability [-0.07%,0.07%] >0.07% <-0.07%  DL Cell Throughput
Remaining KPIs [-3%,3%] >3% <-3%  CQI and DL PDSCH MCS average

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
KPI Comparison:: Traffic Volume and User ( Adaptive Closed Loop MIMO – Batch2 ) CUSTOMER
EXPECTATION

• DL and UL Traffic Volume having similar trend

• Average User Number also having similar trend.

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
KPI Comparison:: Downlink and Uplink Throughput ( Adaptive Closed Loop MIMO – Batch2EXPECTATION
) CUSTOMER

• Cell and User DL Throughput significantly increased around 8.83% due to higher CQI and MCS as per expected
• UL Cell and User Throughput showing stability
NETWORK QUALITY MANAGEMENT | 2017
 #GOBEYOND
KPI Comparison:: Accessibility and Retainability ( Adaptive Closed Loop MIMO- Batch2 ) CUSTOMER
EXPECTATION

• RACH Contention Resolution showing stability

• RRC, ERAB Setup Success & Service Drop Rate showing stability.
NETWORK QUALITY MANAGEMENT | 2017
 #GOBEYOND
KPI Comparison:: Mobility and CSFB ( Adaptive Closed Loop MIMO – batch2 ) CUSTOMER
EXPECTATION

• IFHO and CSFB showing improve but not related to this activity

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
KPI Comparison:: PRB Utilization, CQI & Rank2 Rate ( Adaptive Closed Loop MIMO- batch2EXPECTATION
) CUSTOMER

• DL/UL PRB Utilization showing reduced but not related to this activity
• CQI >7 Distribution improve around 6.35% as expected
• MIMO
NETWORK rank-2
QUALITY are sdecreased
MANAGEMENT around 7% due to PMI reporting
| 2017
 #GOBEYOND
KPI Comparison:: IBLER, MCS and UL Interference ( Adaptive Closed Loop MIMO- Batch2 ) CUSTOMER
EXPECTATION

• DL MCS increased as expected around 11.99%

• DL and UL IBLER are decreased -0.55% and -2.29%
• UL interference show stable trend
NETWORK QUALITY MANAGEMENT | 2017
 #GOBEYOND
Additional KPI:: Related Feature/Parameter ( Fixed Closed Loop MIMO – Batch2 ) CUSTOMER
EXPECTATION

• Transmission Mode distribution switch between TM4 and TM2

• Rank distribution are changing from Open loop rank1 and rank2 into Closed loop rank1 and rank2

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
Summary KPI Comparison Adaptive Closed Loop MIMO Batch #2 - POI CUSTOMER
EXPECTATION

Classes KPI Before (22-23 March ) After (1- 2 March) Delta (%) Formula Remark
Rach Cont. Resol SR (%) 83.28 82.79 -0.49 After-Before maintain
RRC setup SR (%) 99.98 99.98 0.00 After-Before maintain
Accessibility
E-RAB setup SR (%) 99.95 99.94 -0.01 After-Before maintain
Call Setup Success Rate (%) 99.91 99.89 -0.02 After-Before maintain
Retainability Service Drop Rate (%) 0.08 0.08 0.00 After-Before maintain
Mobility Intra-Frequency Handover Out Success Rate (%) 99.81 99.83 0.03 After-Before maintain
CSFB CSFB Preparation Success Rate (%) 99.89 99.90 0.02 After-Before maintain
Cell Downlink Average Throughput 15.98 17.39 8.82 (After-Before)/Before*100 improved
Cell Uplink Average Throughput 1.77 1.92 8.65 (After-Before)/Before*100 improved
Service Integrity
User Downlink Average Throughput 10.32 12.34 19.60 (After-Before)/Before*100 improved
User Uplink Average Throughput 1.68 1.82 8.61 (After-Before)/Before*100 improved
DL Resource Block Utilizing Rate (%) 21.59 20.47 -1.13 After-Before maintain
Utilization
UL Resource Block Utilizing Rate (%) 13.60 12.78 -0.82 After-Before maintain
Average User Number 17212.17 17179.38 -0.19 (After-Before)/Before*100 maintain
User
Maximum User Number 11854.80 11901.20 0.39 (After-Before)/Before*100 maintain
Downlink Traffic Volume 890685.01 925413.32 3.90 (After-Before)/Before*100 improved
Traffic
Uplink Traffic Volume 77495.21 80458.90 3.82 (After-Before)/Before*100 improved
Availability Radio Network Availability Rate (%) 93.67 93.67 0.00 After-Before maintain
Interference Average UL interference -107.21 -106.56 -0.61 (After-Before)/Before*100 maintain
CQI>7 (%) 87.12 91.12 4.00 After-Before improved
DL MCS 13.15 14.43 9.76 (After-Before)/Before*100 improved
Quality
UL MCS 16.25 16.31 0.35 (After-Before)/Before*100 maintain
MIMO Rank2 Ratio (%) 41.23 32.99 -8.24 After-Before degraded
DL IBLER (%) 10.77 11.45 -0.68 After-Before degraded
IBLER
UL IBLER (%) 11.73 13.36 -1.63 After-Before degraded

KPIs Maintain Improved Degrade Significant improvement:

Retainability [-0.03%,0.03% >0.03% <-0.03%  DL User Throughput
Accessibility & Availability [-0.07%,0.07%] >0.07% <-0.07%  DL Cell Throughput
Remaining KPIs [-3%,3%] >3% <-3%  CQI and DL PDSCH MCS average

NETWORK QUALITY MANAGEMENT | 2017

 #GOBEYOND
CUSTOMER
EXPECTATION

Thank You…

NETWORK QUALITY MANAGEMENT | 2017