Mobifone Network performance analysis

1 Network performance benchmarking and analysis

2 Recommended feature's

3 ProACTive program introduction

1 © 2021 Nokia public

Mobifone Network performance analysis

1 Network performance benchmarking and analysis

2 Recommended feature's

3 ProACTive program introduction

2 © 2021 Nokia public

 Network Analysis Insights Logo
of
Operator

Mobifone Vietnam Network Analysis

3rd June, 2022

SW Sales, MN RBC, APJ

3 © 2021 Nokia public

 1 Mobifone Mobile Network Performance Benchmark

Mobifone Network Analytics using OSS stas

2
(Nokia Footprint)

3 Summary and Recommendations

4 © 2021 Nokia public

Mobifone Vietnam - Network Analysis
1. Mobifone is leading in Downlink and Uplink performance in
North, but running behind competition in South and
Middle (Source : Tutela)
2. Latency improvement needed in all regions.
3. Mobifone is leading in user experience across all
categories in North, but ranking no.2-3 in other regions
4. SW Features activation has improved. Should continue to
improve performance further
5. SW release upgrade to latest release 22R1 is mandatory
to implement Dynamic spectrum sharing with L21 and
other good features for power savings
6. Sector upgrade and 4x4MIMO upgrade recommended for
additional capacity
5 © 2021 Nokia public
 1 Mobifone Mobile Network Performance Benchmark

Mobifone Network Analytics using OSS stats

2
(Nokia Footprint)

3 Summary and Recommendations

6 © 2021 Nokia public

Nokia continuing global LTE performance leadership
Delivering the highest average data speeds across 37 countries and 236 cities

Average download speed (Mbps)

1 July 2021 19 November 2021

Average 4G downlink speed (Mbps) Average 4G uplink speed (Mbps) Average 4G latency (ms)
Nokia
32 31 14 42 Supplier A
38.7
37.7
12.2
38 35.9
Supplier B
28 11.9
28 12 11.6
27 34 Nokia analysis based
on crowdsourced
30
24 10 data from Tutela
26 Technologies, Ltd.

20 8 22

7 © 2021 Nokia Confidential

 Summary of KPI benchmarking : Competitors lead in key KPIs
Nationwide

Mobifone has best packet loss. Focus on improving DL & UL throughput and Latency
Sources: Nokia analysis based on data from Tutela (Apr 2022)

▪ Leading in performance for Packet

Loss

▪ Mobifone lagging competitors in

DL Th’put performance & CQI

9 © 2021 Nokia public

 DL user throughput consistently no.3 for Mobifone. Likely from rising data usage
Nationwide

Latency has improved in the past, but still behind competition

Sources: Nokia analysis based on data from Tutela (Apr 2022)

10 © 2021 Nokia public

 Mobifone leads in Uplink, Packet Loss and Jitter. Close#2 in Downlink
North - Nokia

Latency requires improvement

11 © 2021 Nokia public

North - Nokia
Mobifone leading in DL&UL throughput in Hanoi, Lao Cai and Dian Bian provinces
Lagging in the rest. Improvements needed
Sources: Nokia analysis based on data from Tutela (Apr 2022)

12 © 2021 Nokia public

North - Nokia

Sources: Nokia analysis based on data from Tutela (Apr 2022)

65%

65%

13 © 2021 Nokia public

 Tutela Experience Index - Estimating Quality of Experience using crowdsourced data
Measure of how likely mobile networks are meeting subscriber expectations
Tutela Experience Index

Core Targeted
(pervasive apps and services used by most subscribers) (common apps and services used by large minorities of
users)
Weightings 20% 20% 20% 10% 10% 10% 10%

Video Web Connected Content Network Video

VoIP
Streaming Browsing Apps Sharing Gaming Calling
KPIs impacting experience score

d/l throughput

u/l throughput

latency

packet loss

jitter

14 © 2021 Nokia Confidential

North - Nokia
Mobifone leading in user experience benchmark due to strong DL & UL
throughput and Packet Loss performance
Sources: Nokia analysis based on data from Tutela (Apr 2022)

15 © 2021 Nokia public

 Mobifone lagging in all KPIs
South - Nokia

High traffic could potentially be contributing to it

Sources: Nokia analysis based on data from Tutela (Apr 2022)

16 © 2021 Nokia public

 Mobifone DL performance is poorer in all provinces except in Binh Thuan
South - Nokia

Uplink poorest in all provinces – activation of uplink features and optimization will help
Sources: Nokia analysis based on data from Tutela (Apr 2022)

17 © 2021 Nokia public

 Mobifone has worse than competition latency in all provinces
South - Nokia

41% of latency samples are between 30 to 50ms. Ideal is closer to 30ms

Sources: Nokia analysis based on data from Tutela (Apr 2022)

41%

41%

18 © 2021 Nokia public

 2% of the Packet Loss samples are >2% compared to Vinaphone
South - Nokia

Not a big problem, but improvement will ensure leading rank in Video Streaming application
Sources: Nokia analysis based on data from Tutela (Apr 2022)

19 © 2021 Nokia public

South - Nokia

Mobifone ranks No.2 in user experience in most applications categories

Poorer KPI performance across the board leading to loss of leadership in user experience
Sources: Nokia analysis based on data from Tutela (Apr 2022)

20 © 2021 Nokia public

Middle - Nokia
Mobifone lagging behind competition in most KPIs
Improvements required
Sources: Nokia analysis based on data from Tutela (Apr 2022)

21 © 2021 Nokia public

Middle - Nokia
Mobifone DL performance better than Viettel but worse than Vinaphone in most provinces

Sources: Nokia analysis based on data from Tutela (Apr 2022)

22 © 2021 Nokia public

Middle - Nokia
Mobifone having large number of Latency samples between 30-50 ms
Latency optimization strongly recommended to improve user experience
Sources: Nokia analysis based on data from Tutela (Apr 2022)

87%

23 © 2021 Nokia public

Middle - Nokia
5% of Mobifone samples having more than 2% packet loss
4.7% of high packet loss samples due to poor coverage quality, 20% due to low DL thrpt and rest
possibly due to transport / cre network issues

Sources: Nokia analysis based on data from Tutela (Apr 2022)

24 © 2021 Nokia public

Middle - Nokia
Mobifone ranks no.2 in user experience in most applications categories

Sources: Nokia analysis based on data from Tutela (Apr 2022)

25 © 2021 Nokia public

Journey needs to start in order to finish
• 5G starting to grow

S-curve of Technology evolution

• South East asia yet
to see a ramp up
• Devices, ecosystem,
coverage …

• Very early days for

Myanmar, but
Korea, US, sooner we start
Western Europe sooner we will get
South East Asia to growth phase
• Learnings, use cases
• 4G still your cashcow

26 © 2022 Nokia
5G UE penetration is in double digits and rising fast in Vietnam

• 3G UEs contribute much less record samples, %ages based on 4G+5G records only
• 5G UE penetration could adjusted accordingly (typically 80-90% of values shown)
27 © 2021 Nokia public
 1 Mobifone Mobile Network Performance Benchmark

Mobifone Network Analytics using OSS data

2
(Nokia Footprint)

3 Summary and Recommendations

28 © 2021 Nokia public

 Network level KPI benchmarking and
trend (April 2022 data from OSS)

29 © 2021 Nokia public

Nationwide
Overall data traffic increase by 17% in last 12 months in Nokia footprint
Driven by LTE that grew 22% over the year. LTE data contribution is around 94%

Data 3G
traffic -48%
05 Apr 2021 LTE
To
28 Mar 2022 +22%

Total Voice traffic decreased by -35%

Voice 2G
traffic -32%
05 Apr 2021 3G
To
28 Mar 2022 -43%

30 © 2022 Nokia Confidential

 NetAct

Avg daily data usage per cell grew 35% over last 12 months from 41 GB TO 56 GB Nokia
footprint
Avg Daily DL PRB utilization increased almost 41% during the period, indicating rising load in cells

DL Data/cell/day ranking as of Mar'22 DL Data/cell/day increasing as of Q2’21

DL Data/cell/day
+35%
(GB)

• Moderate compared to
other Nokia networks
globally, but rising fast 56 (40)

• Covid-19 impact in Q1.

DL PRB ranking as of Mar'22 DL PRB decreased since Q2’21

Avg daily DL PRB

+41%
• Low compared to other
Nokia networks globally

• Increasing since Q2’21

17 (62)

31 © 2022 Nokia Confidential

 NetAct

Average Connected users per cell still low, but rising fast Nokia
footprint
Maintained moderate network quality level
Connected UEs LTE Connected UEs ranking as of Mar'22 Connected UEs increasing since Q2’21

• Low compared to
other Nokia
networks globally +40%
and Increasing since
Q2’21.
• Nokia Active cell
decreased since 14 (54)
Nov’21.

Ave CQI ranking as of Mar'22 CQI has slightly decreased since Q2’21

CQI
• CQI at moderate
level 10 (45)

• Potential to improve

32 © 2022 Nokia Confidential

 NetAct

DL end user Throughput for QCI9 (LTE_5358) and Spectral efficiency showing slight Nokia
footprint
downward trend

Confidential
33 © 2022 Nokia Confidential
Steps to ensure best end-user performance
Leverage contracted software content for improved performance and adopt a focused
investment strategy in building capacity

1 2 3
Optimisation Software release Software features
Improve performance Upgrade to latest release Active performance enhancing
from existing assets, for improved features for DL/UL, VoLTE,
especially if quality is poor performance and access load balancing, interworking,
(interference), utilise SON to latest features Carrier aggregation etc.

4 5 6
Capacity expansions Spectrum / refarming Add new sites
Bandwidth upgrade, MIMO, Add or refarm spectrum Coverage or densification
Add Sector to existing sites, DSS e.g. Small Cells or new
(dynamic spectrum macro sites
sharing with L21)

34 © 2021 Nokia public

3% of cells have moderately low CQI<7, indicating inferior network quality
4360 cells; 28% at CQI<9. Optimization needed for poor-quality cells (as of May’22)
North South
CQI in decreasing trend 4360 cells have CQI <9. 489 cells CQI below 7 10926 cells have CQI <9. 1874 cells CQI below 7
Total #cells, 15k+ Total #cells, 23k+
Avg CQI
11.5

11 4360 10926
10.5

10

9.5

9
Jul 2021
Jun 2021
May 2021

Apr 2022

May 2022
Aug 2021

Sep 2021

Nov 2021

Dec 2021

Feb 2022
Mar 2022
Jan 2022
Oct 2021

489 1874

<7 <9 <7 <9

L1800 475 4298 L1800 1472 9343
L2100 14 62 L2100 402 1583
- - - -
- - - -
- - - -

489 4360 1874 10926

35 © 2021 Nokia public

Confidential
 SW value proposition

36 © 2021 Nokia public

Adoption of new SW release and features lead to improved
performance and user experience
DL and UL throughput as well as Latency improved Strong correlation between SW features use and
with new release upgrade spectral efficiency*

Performance improvement with SW upgrade* Spectral efficiency vs feature activation

30 60
56 2.2
49

LTE DL Spectral Efficiency (bit/s/hz)

25 28 50
2
Throughput (Mbps)

20 40
22

Latency (ms)
1.8
29
15 30

14 1.6
10 20
11
9 1.4
5 8 10

0 0 1.2
SRAN19 or earlier SRAN20A/B/C SRAN21A
1
20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 70%
DL T'put (Mbps) UL T'put (Mbps) Latency (ms)
Feature activation index

* Based on data from 68 Nokia LTE customers with big installed base. Other Study across 55 networks for highly loaded cells (PRB >70%)
events such as site expansions, upgrade of transport networks, change of traffic
load etc. will also impact NW performance
* Spectral efficiency may vary based on RF conditions (CQI level) and available
bandwidth (higher bandwidth will in general provide better spectral efficiency).

37 © Nokia 2019
Field results from recent SW upgrades: + 4-5% downlink throughput
Basic SW and enhanced algorithms improve performance (no new features activated)

Release upgrade example from Latin America Release upgrade example from Europe
Release
upgrade DL throughput Mbps
+4.3%
DL throughput Mbps
+4.7%

29.9 31.2
22.3 23.4

* Measured on
BTS level one
week before and SRAN20B SRAN21B SRAN20C 22R1
one week after
upgrade using • Upgrade of 2.2K BTS in March 2022 • Upgrade of 728 BTS in March 2022
Tutela • BTS level downlink throughput* increased by 4.7% • BTS level downlink throughput* increased by 4.3%

38 © 2022 Nokia Confidential

Field results from LTE and 5G performance features
Four examples of improved uplink throughput

LTE LTE 5G 5G
SW Traffic driven UL CA SCell UL 256QAM Long PUCCH for FR1 TDD 256 QAM for PUSCH
features configuration
(SRAN20A: LTE4040) (SRAN19B: LTE4581) (5G21A: CB006064) (5G21B: 5G001118)

UL throughput Mbps UL throughput Mbps UL throughput Mbps UL throughput Mbps

+28% +30% +10% +23%

69.4 65 85 74
54.4 91.1
74.4

67

Off On Off On
Off On Off On

39 © 2022 Nokia Confidential

 After features activation user experience and resource
utilization improved on networkwide basis
UL user throughput improved 35% since April DL user throughput improved 8% since June
• Better user
Possible Covid
impact
+12% experience
+35%

PCell swap UL CA
after features
activated activated DL CA for 12
layers activated activation

Daily GB/cell rose 30%, load only 21% … And Spectral Efficiency improved 6%

GB %
• More efficient
+30% +6%
resource
utilization
+21%

40 © 2022 Nokia Case Study: Oceania customer

Confidential

Enhance the customer experience and sustainability

22R1 Highlights – improved customer value with SW, HW and Services

• Unleashed download speeds with • ABIO concurrent 4G+5G mode • Uplink closed loop power control • Dynamic Bandwidth Part
2CC (TDD+TDD) and 3CC CA • ABIO capacity doubled for NR • DSS Superfast coordination Adaptation to extend device
• mmWave (FR2) extended cell battery life
range up to 2.3km • As side effect, save global CO2
• Planning and maintenance effort emissions by less frequent
reduced with new ANR features phone charging

Unleashed download speeds Reduced customer TCO and Enhanced end-user Improve user satisfaction
with 2CC and 3CC CA improved sustainability experience and reduced churn with better mobile phone
battery life
(2

>3Gbps DL >20m€ +5% UL +40%

(1

1) Model network, cumulative effort reduction, over 5 years. 2) With low bandwidth services, total daily impact naturally less
41 ©©
Nokia
20212021
Nokia
Increased Customer Value with SRAN22R1 – expected to provide benefits up to
12 MUSD for MobiFone

Customer Value

SRAN21B Release, Basic

CAPEX
Content (BSW)
Savings &
TCO
Improve-
ment
High Value Features for radio
resource mgmt., capacity, new
service, operability etc. …

NEW
REVENUE
Site Solutions : new HW STREAM
capability, configurations

42 © Nokia 2019
SW Release upgrade is about much more than just corrections
SW subscription is the best way to unlock value from SW Releases
Fixed annual fee 1
$
• Good for mature network
Corrections
• Not optimal for growing network
Stable
SW

RAN
Evolution
(ORAN/vRAN)
Y1 Y2 Y3 Y4 Y5
SW Release
Performance
Upgrade improvement 2
Variable annual fee
$
• Linked to cell#, users, Mbps, Mhz
Interop-
• Good for growing network
erability

New HW /
New Config /topology
functionality support

43 © 2021 Nokia Y1 Y2 Y3 Y4 Y5
SW features activation has improved over last 1 year. Continue activating more..
Upgrade to 22R1 is a mandatory next step for 2nd carrier activation with DSS
Feature Benefit Mobifone Mobifone
Feature Name Release
id Category (Jul'21) (Apr'22)
Software Activation level (4G) LTE1332 Data Perform. DL Downlink carrier aggregation - 40 MHz RL60 7% 3%
LTE1092 Data Perform. UL Uplink carrier aggregation - 2 CC FDD-LTE 16A 0% 0%
100 Data Performance
LTE1819 Data Session Profiling FDD-LTE 16
DL 3%
80 LTE1987 Data Perform. DL DL Adaptive CL SU MIMO (4x4) FDD-LTE 16 11% 12%
LTE2275 Data Perform. UL PCell swap FDD-LTE 16 0%
60 LTE4054 Data Perform. DL Dynamic Power Back-off FDD-LTE 18A 0% 0%
LTE4064 Data Perform. DL 256 QAM enhancements FDD-LTE 18A 3% 3%
40 LTE2664 Data Perform. UL Load Based PUCCH Region FDD-LTE 17A 0% 18%
Data Perform. UL 256QAM in UL FDD-LTE 19B 0%
20 LTE4961 Data Perform. DL Flexible CA management FDD-LTE 19A 2% 0%
LTE1170 Load Balancing Inter eNode B IF Load Balancing RL50 0% 15%
LTE1387 Load Balancing for AirScale SM for EnergyEfficiency RL40 0% 15%
0
LTE1531 Load Balancing Inter-frequency load balancing extension RL60 0% 15%
Load Balancing Dedicat Idle Mode Mobil Prio Support FDD-LTE 16 14%
LTE487 Load Balancing Idle mode mobility load balancing RL50 0% 15%
Based on Apr’22 data Load Balancing CA Aware Idle Mode Load Balancing FDD-LTE 16A 19%
LTE11 VoLTE Robust header compression RL20 0%
LTE3723 VoLTE VoLTE Call Steering Enhancement SRAN 20B 0%
Load based Power Saving for multi-layer
• Increase 4x4/4x2 MIMO penetration for high load sites. LTE1103 Energy
networks
FDD-LTE 15A
0% 0%
Load based Power Saving with Tx path switching
LTE1203 Energy FDD-LTE 16
• Prioritise 3G-4G DSS implementation. After second off 0% 0%
LTE1891 Energy eNode B power saving - Micro DTX FDD-LTE 16 73% 67%
carrier rollout – CA, Load balancing features LTE3036 Energy Micro DTX extension FDD-LTE 18A 44% 55%
Micro discontinuous transmission (µDTX) for
• Activate Power savings features to offset rising energy CB006047 Energy
Energy Efficiency SRAN enhancements
SRAN 21A
100%
costs from increasing traffic SR002546 SRAN DSS Dynamic Spectrum Sharing WCDMA+LTE, phase 2 SRAN 20A 0%
LTE2430 VoLTE QCI1 Triggered Inactivity Timer LTE16 0%
• Plan for VoLTE launch to gradually take over 3G voice LTE2098 VoLTE VoLTE Uplink Coverage Boosting LTE18 0%
LTE5731 Performance TCP Service Differentiation LTE18A 1%
44 © 2021 Nokia public
Source: OSS statistics in Nokia footprint
Cells with 4x4 MIMO have 22% better spectral efficiency and 20% higher user throughput
and carry 31% higher traffic at comparable cell load
Upgrade the rest of 2x2 MIMO to 4X4 MIMO to increase capacity

+22% +20%

45 © 2021 Nokia Confidential

25% site level improvement with RAN energy efficiency features

Embedded power meter pre-installed all AirScale radios and baseband

RF Paths Micro Cuts RF Paths Switch Off BB Capacity Switch Off

Micro DTX Cell shut down (m)MIMO muting Baseband dynamic

power management

Micro DTX scheduler

optimization
10% - 15% reduction in 10% - 30% reduction in
electricity consumption electricity consumption
(Deep) Sleep Mode

10% reduction in
10% - 20% reduction in electricity consumption
electricity consumption of base band
(>40% in no load (5G) Additional 12% reduction in IDLE electricity consumption
situation)
46 © 2021 Nokia Nokia internal use
 Energy power saving potential for Mobifone of 1.1m Euro per year
North - Nokia

Site level analysis based on current configurations, network load and software activation potential
Mobifone traffic load in network over 24 hour period on site level Site level power consumption and saving potential
(Hours) (kWh/day)
24
80

20

60
Total daily energy consumption
(kWh) before activation of features
15

Site level savings (daily kWh) based on

40
recommended features
10

20
5

0
0
(Sites) (Sites)

Activation Possible Saving Potential

Recommendations Level (%) Activation (kWh) Estimated Power Savings
LTE1203 Load-based Power Saving with Tx path switching off 1.68 4813 3.42M
Total sites in Network: 4939 Saving in kWh (Yearly) 3.8 Mio
LTE1103 Load-based Power Saving for Multi-layer 0 165 0.38M
LTE2508 BTS Embedded Power Meter for EE Monitoring 1.09 4885 Total Cells in Network: 15548 Energy cost per kWh 0.30 Euro (*)
LTE3036 Micro DTX extensions
# of applicable sites: 4755 Total saving (yearly): 1.14 mEuro
Energy Saving Implementation Solution using EdenNet

* Including DG powered sites (est.)

47 © 2021 Nokia Confidential
 Energy power saving potential for Mobifone of 1.3m Euro per year
South - Nokia

Site level analysis based on current configurations, network load and software activation potential
Mobifone traffic load in network over 24 hour period on site level Site level power consumption and saving potential
(Hours) (kWh/day)
24
80

20

60
Total daily energy consumption
(kWh) before activation of features
15

Site level savings (daily kWh) based on

40
recommended features
10

20
5

0
0
(Sites) (Sites)

Activation Possible Saving Potential

Recommendations Level (%) Activation (kWh) Estimated Power Savings
LTE1203 Load-based Power Saving with Tx path switching off 0.02 6374 2.85M
Total sites in Network: 6534 Saving in kWh (Yearly) 4.41 Mio
LTE1103 Load-based Power Saving for Multi-layer 0.02 1104 1.56M
LTE2508 BTS Embedded Power Meter for EE Monitoring 70.97 1897 Total Cells in Network: 23809 Energy cost per kWh 0.3 Euro (*)
LTE3036 Micro DTX extensions Already activated
# of applicable sites: 6165 Total saving (yearly): 1.3 mEuro
Energy Saving Implementation Solution using EdenNet

* Including DG powered sites (est.)

48 © 2021 Nokia Confidential
 Nokia ESM Module in EdenNet SON enhances
power savings potential further Legacy ESM
100 aggregated traffic
AI / ML based algorithm enhances Activation Period

Load
90 Power Saving Period load in Power Group
80
70
60
50
40
Functional Requirement: 30
20 *Switch-on threshold
• Prior to this feature, operators configure Energy Savings (ES) activation periods in a
manual manner. 10 *Switch-off threshold
0
• Since each cell’s load period is different, therefore Machine Learning (ML) helps to

0:00
1:00
2:00
3:00
4:00
5:00
6:00
7:00
8:00
9:00

0:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
predict the load of every cell, and configure optimal activation periods to maximize
Energy savings from cell switch-off.
ML-enhanced ESM: Additional power savings activation
periods based on machine learning of cell load.
100
ML enhanced Power Saving Periods

Load
90
80
Customer Benefit: 70
60
✓ ML potentially increases the period of cell-switch-off by predicting optimal time-
50
periods of low load in the power cell group
40
✓ ML prediction per power savings group accurately configures activation periods per 30
the load characteristics experienced by each set of co-sector cells 20
10
0

0:00
1:00
2:00
3:00
4:00
5:00
6:00
7:00
8:00
9:00

0:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
17:00
18:00
19:00
20:00
21:00
22:00
23:00
49 © 2021 Nokia Nokia internal use
*) The ML capability is license controlled
 Cell level capacity analysis
(North region)

50 © 2021 Nokia public

 NetAct
Mobifone has 4.9k LTE BTS from Nokia with a total of 15.5k cells
North - Nokia
Nokia
across 2 bands footprint

Band Bandwidth (MHz) Cells

- 1800 20/15/10 14961/31/38 (15030)
- 2100 5 509 (509)

509

51 © 2021 Nokia Confidential

North - Nokia

L1800(B3)
(20 MHz)

53 © 2021 Nokia Confidential

North - Nokia

L2100(B1)
(5 MHz)

54 © 2021 Nokia Confidential

North - Nokia
Sector level HW capacity congestion analysis & recommendations
Congested Cells Recommendations

Sector Level Band Combination # of Sectors L1800_20/ HW Expansion Path Load

L1800_20 L2100_5 L1800_10
L2100_5 Recommendations Balance

L2100_5 16 2 2

L1800_15 26
Add L21 layer, 4x4
L1800_20 14264 471 471
MIMO upgrade
L1800_10 33 4 BW upgrade, Add L21 4

L2100_5/L1800_15 3
L18 MIMO upgrade,
L1800_20/L2100_5 490 21 9 7 37
L21 BW upgrade
L1800_20/L1800_15 2

L1800_20/L1800_10 5 1 1

Total # of Sectors 14839

Total Congested Sector 515 492 11 5 7 514 1

Thresholds: DL PRB>=80% and End user T’put<=5 Mbps

56 © 2021 Nokia Confidential

North - Nokia
Sector level HW capacity congestion analysis & recommendations - Projection
Considering 35% growth in next 12 months
Congested Cells Recommendations

Sector Level Band Combination # of Sectors L1800_20/ HW Expansion Path Load

L1800_20 L2100_5 L1800_10
L2100_5 Recommendations Balance

L21 BW upgrade, Add

L2100_5 16 5 5
L18 layer
L1800_15 26
Add L21 layer, 4x4
L1800_20 14264 900 900
MIMO upgrade
L1800_10 33 9 BW upgrade, Add L21 9

L2100_5/L1800_15 3
L18 MIMO upgrade,
L1800_20/L2100_5 490 33 23 23 71 8
L21 BW upgrade
L1800_20/L1800_15 2

L1800_20/L1800_10 5 1 1

Total # of Sectors 14839

Total Congested Sector 994 933 28 10 23 985 9

Thresholds: DL PRB>=59.26 (80%/1.35) and End user T’put<=6.75 (5*1.35 Mbps)

57 © 2021 Nokia Confidential

 Cell level capacity analysis
(South region)

58 © 2021 Nokia public

South - Nokia NetAct

Mobifone has 6.5k LTE BTS from Nokia with a total of 23.7k cells across 2 bands Nokia
footprint

Band Bandwidth (MHz) Cells

- 1800 20/15/5 19099/1335/7 (20441)
- 2100 5 3291 (3291)

59 © 2021 Nokia Confidential

South - Nokia

L1800(B3)
(20/15 MHz)

L18 (20Mhz) L18 (15Mhz)

60 © 2021 Nokia Confidential

South - Nokia

L2100(B1)
(5 MHz)

61 © 2021 Nokia Confidential

South - Nokia
Sector level HW capacity congestion analysis & recommendations
Congested Cells Recommendations

Sector Level Band Combination # of Sectors L1800_20/ HW Expansion Path Load

L1800_5 L1800_15 L1800_20 L2100_5 L1800_10
L2100_5 Recommendations Balance

L1800_5 7 7 BW upgrade 7

L2100_5 24 3 BW Upgrade 3

L1800_15 1323 10 BW upgrade, Add L21 10

Add L21, 4x4MIMO
L1800_20 15640 1386 1386
upgrade in L18
L1800_10 7 1 BW upgrade 1

L2100_5/L1800_15 6 3 1 BW upgrade 4
L21 BW upg., L18
L1800_20/L2100_5 3259 559 45 332 923 13
MIMO upg. Smallcells
L1800_20/L1800_15 2

Total # of Sectors 20268

Total Congested Sector 2347 7 13 1945 49 1 332 2334 13

Thresholds: DL PRB>=80% and End user T’put<=5 Mbps

63 © 2021 Nokia Confidential

South - Nokia
Sector level HW capacity congestion analysis & recommendations - Projection
Considering 35% growth in next 12 months
Congested Cells Recommendations

Sector Level Band Combination # of Sectors

L1800_20/ HW Expansion Path Load
L1800_5 L1800_15 L1800_20 L2100_5 L1800_10
L2100_5 Recommendations Balance

L1800_5 7 7 0 0 0 0 0 BW upgrade 7

L2100_5 24 0 0 0 3 0 0 BW Upgrade 4
BW upgrade, Add
L1800_15 1323 0 10 0 0 0 0 16
L21
Add L21, 4x4MIMO
L1800_20 15640 0 0 1386 0 0 0 2200
upgrade in L18
L1800_10 7 0 0 0 0 1 0 BW upgrade 2

L2100_5/L1800_15 6 0 3 0 1 0 0 BW upgrade 4
L21 BW upg., L18
L1800_20/L2100_5 3259 0 0 559 45 0 332 MIMO upg. 1395 77
Smallcells
L1800_20/L1800_15 2 0 0 0 0 0 0

Total # of Sectors 20268

Total Congested Sector
3705 7 19 2879 105 2 693 3628 77
Thresholds: DL PRB>=59.26 (80%/1.35) and End user T’put<=6.75 (5*1.35 Mbps)

64 © 2021 Nokia Confidential

 1 Mobifone Mobile Network Performance Benchmark

2 Mobifone Network Analytics (Nokia Footprint North)

3 Summary and Recommendations

65 © 2021 Nokia public

Summary and Recommendations
1. Mobifone is leading in Downlink and Uplink performance in North, but running behind
competition in South and Middle (Source : Tutela). Latency improvement needed in all regions.
2. Mobifone is leading in user experience across all categories in North, but ranking no.2-3 in
other regions
3. SW Features activation has improved. Should continue to improve performance further
4. SW release upgrade to latest release 22R1 is mandatory to implement Dynamic spectrum
sharing with L21 and other good features for power savings
5. Key Recommendations –
• Deploy L21 with DSS. Add sectors or 4x4MIMO upgrades selectively as needed.
• Optimization needed for cells below CQI 7
• Upgrade SW Release to 22R1

LTE1332 Downlink carrier aggregation - 40 MHz LTE11 Robust header compression Load based Power Saving for multi-layer
LTE1103
LTE1092 Uplink carrier aggregation - 2 CC LTE3723 VoLTE Call Steering Enhancement networks
LTE1819 Data Session Profiling SR002546 Dynamic Spectrum Sharing WCDMA+LTE, phase 2 Load based Power Saving with Tx path switching
LTE1203
QCI1 Triggered Inactivity Timer
off
LTE1987 DL Adaptive CL SU MIMO (4x4) LTE2430
LTE2275 PCell swap LTE2098 VoLTE Uplink Coverage Boosting
LTE4054 Dynamic Power Back-off LTE4961 Flexible CA management
LTE4581 256 QAM in Uplink LTE5731 TCP Service Differentiation
LTE2664 Load Based PUCCH Region

66 © 2021 Nokia public

Mobifone Network performance analysis

1 Network performance benchmarking and analysis

2 Recommended feature's

3 ProACTive program introduction

67 © 2021 Nokia public

Mobifone
Feature recommendations

June 2022
Jari Malinen

68 © 2022 Nokia Confidential

LTE features to enhance user experience
Performance and Energy Efficiency are on everyone’s agenda
Mobifone Customer
Benefit Category Feature id Feature Name Release
VN use
LTE1819 Data Session Profiling FDD-LTE 16 3% ●●●●● *
LTE2276 Measurement Based SCell Selection FDD-LTE 16A 0% ●●●
LTE4054 Dynamic Power Back-off FDD-LTE 18A 0% ●●●● *
Data Performance DL
LTE4064 256 QAM enhancements FDD-LTE 18A 3% ●●●●●
LTE4961,LTE5426,
Flexible CA management FDD-LTE 19A 0%
LTE5427,LTE5673 ●●●●
LTE1092 Uplink carrier aggregation - 2 CC FDD-LTE 16A 0% ●●●●●
LTE4050 UL traffic driven UL CA SCell conf SRAN 20A ●●
Data Performance UL
LTE4581 256QAM in UL FDD-LTE 19B 0% ●●●●
CB007440 Split ULPC parameters for 256QAM SRAN 21B ●
Performance LTE5731 TCP Service Differentiation LTE18A 1% ●●●● *
LTE4066 Low Latency Fast UL Grants FDD-LTE 18A 0% ●●●●●
Latency
LTE815 Proactive UL Resource Assignment RL10 1% ●●●●●
LTE2050 Load Triggered Idle Mode Load Balancing RL70 0% ●●●●●
Load Balancing
LTE2051 Measurement based Idle Mode Load Balancing FDD-LTE 15A 0% ●●●●
LTE1103 Load based Power Saving for multi-layer networks FDD-LTE 15A 0% ●●●● *
CB007416 Load based Cell Switch Off/On for Power Saving SRAN 21B ●●
SR000951 BTS Embedded Power Meter for Energy Efficiency Monitoring SRAN 16.10 59 % ●●●●●
Energy
LTE1203 Load based Power Saving with Tx path switching off FDD-LTE 16 0% ●● *
SR003080 Power saving with online TX path switching off SRAN 21B ●
LTE3036 Micro DTX extension FDD-LTE 18A 55 % ●●●●
* Recommended 2021

69 © 2022 Nokia
LTE features for voice quality and optimal frequency use
VoLTE, DSS and operability
Mobifone Customer
Benefit Category Feature id Feature Name Release
VN use
Operability SR001155 BTS Rx-Sniffing SRAN 18A 0% ●●●●
SR001933 Automatic sleeping cell detection and recovery SRAN 18A 0% ●●
Serviceability SR002240 RX Power and traffic based sleeping LTE cell detection SRAN 19B 0% ●●
SR002270 KPI triggered alarm and log collection in BTS SRAN 20B 0% ●
SR001783,
SR002532, SRAN Dynamic Spectrum Sharing GSM+LTE SRAN 19
SRAN DSS SR002596 ● *
SR002167,
Dynamic Spectrum Sharing WCDMA+LTE SRAN 19B
SR002546 ● *
LTE11 Robust header compression RL20 0% ●●●●● *
LTE2774 Source rate UL grant assign. VoLTE FDD-LTE 18A 0% ●●● *
LTE3723 VoLTE Call Steering Enhancement SRAN 20B 0% ●
LTE1406 Extended VoLTE Talk Time RL70 0% ●●●● *
LTE1942 Dedicat VoLTE inter/target freq list LTE15A 0% ●
VoLTE
LTE2430 QCI1 Triggered Inactivity Timer LTE16 0% ●●●●● *
LTE64 Service Based Handover Thresholds LTE16 0% ●●●●●
LTE2098 VoLTE Uplink Coverage Boosting LTE18 0% ●●●●● *
LTE3692 VoLTE Robustness Enhancements LTE18A 0% ●●●●● *
LTE2832 SRVCC Due to Admission Cntrl Reject. LTE19 0% ● *

* Recommended 2021

70 © 2022 Nokia
 Performance / Latency

73 © 2022 Nokia Confidential

 LTE16
LTE1819
Data Session Profiling
• The main benefit of LTE1819 is to save battery and improve browsing delays.
• Data Session Profiling adapts the UE inactivity timer based on UE data session short term history
• The RRC connection is released when there is no uplink and downlink data transmitted during the UE inactivity timer duration.
LTE1819 optimizes the timer value, making it UE-specific and adapting the value to the UE short-term data activity profile. A
short inactivity timer, which is equal to the value of the inactivityTimer LNCEL parameter, is used by default.

Performance Summary: Inactivity Timer Change

➢ DL and UL payload are stable.
➢ Throughput increase is not related to trial, rather this
increase is from MAXBITRATE change. However there is no
negative impact on throughput from this trial.
➢ RRC setup SR, RRC drop rate, Cont based RACH SR, ERAB
drop rate has improved significantly.
➢ Average RRC users has reduced and paging attempts has
increased, paging discards were increased and necessary
soft action will be performed after accessing panging load.
Performance Summary: Data Session Profiling
➢ RACH setup SR has improved by 0.4%.
➢ ERAB setup SR improved by 0.02%.
➢ ERAB drop rate improved by 0.01%.
➢ Payload and throughput are stable.
➢ RRC users decreased and paging attempts has increased.
➢ All other KPIs are stable.

74 © 2022 Nokia
 LTE16
LTE1819
Field result from APJ -10.75%
PRE (9-11 Nov) POST (16-18 Nov) % Gain/ Loss
KPIs unit
GC FDD TDD GC FDD TDD GC FDD TDD
RRC Attempts # 4958 1869 3088 4309 1747 2563 -13.07% -6.55% -17.02%
ERAB Attempts # 4399 1378 3021 3637 1213 2424 -17.32% -11.95% -19.77%
RRC_PAGING_REQUESTS # 4330475 1424840 2905635 3865092 1398629 2466463 -10.75% -1.84% -15.11%
DL User throughput (Average) Mbps 4.52 3.25 6.00 4.73 3.56 5.87 4.47% 9.52% -2.16%
RRC Connected UE (Average) # 56.27 21.02 35.25 57.07 19.88 37.19 1.42% -5.40% 5.49%
DL Cell throughput (Average) Mbps 14.69 10.45 18.28 14.87 10.63 18.26 1.23% 1.79% -0.13%
UL Cell throughput (Average) Mbps 2.76 1.51 5.57 2.80 1.49 5.54 1.44% -1.46% -0.49%
DL Data Volume (Average) GB 3.27 2.26 4.14 3.28 2.22 4.20 0.50% -1.60% 1.48%
UL Data Volume (Average) GB 0.31 0.25 0.36 0.31 0.24 0.37 -1.30% -5.98% 1.39%
PRB DL utilization (Average) % 50.71 48.13 52.92 49.97 46.04 53.36 -1.46% -4.34% 0.83%
PRB UL utilization (Average) % 41.27 33.00 48.41 41.34 32.19 49.22 0.16% -2.46% 1.66%
RRC Setup Success Rate % 99.82 99.81 99.84 99.85 99.80 99.89 0.03% 0.00% 0.05%
ERAB Setup Success Rate % 99.75 99.72 99.78 99.72 99.68 99.75 -0.04% -0.04% -0.03%
Active Drops per QCI % 0.288 0.475 0.127 0.131 0.198 0.075 -54.39% -58.42% -41.01% -13.07%
Intra Frequency HOSR % 99.17 98.87 99.43 99.21 98.84 99.53 0.04% -0.03% 0.10%
RSSI PUCCH (Average) dBm -104.50 -104.17 -104.77 -104.51 -104.17 -104.81 0.01% 0.00% 0.03%
RSSI PUSCH (Average) dBm -105.17 -106.08 -104.34 -105.10 -105.85 -104.46 -0.06% -0.23% 0.12%
CQI (Average) # 10.02 9.36 10.58 10.06 9.48 10.55 0.41% 1.31% -0.28%

1. Stable Overall KPI performance after feature activation. No impact on Globe

Basic KPIs (Accessibility and Retainability).
2. -10.75% Reduced RRC Paging Request Golden Cluster level
3. -13.07% Reduced RRC Attempts Golden Cluster level -17.32%
4. -17.32% Reduced ERAB Attempts Golden Cluster level
5. +4.47% Improved DL User throughput Golden Cluster level
6. +1.23% Improved DL Cell throughput Golden cluster level
7. Long and Short Inactivity timer UE release due to Data Session Profiling
counter pegging observed
8. Decrease on RRC attempts, ERAB attempts & Paging Request indicates a
reduction on network signaling load
75 © 2022 Nokia
 LTA18A

DL 256QAM enhancement - Dynamic Power Backoff LTE4054

Dynamical power backoff per TTI on a per need bases

allowed EVM

Power Backoff needed

3GPP, EVM limit 3.5% for

256QAM
• Error Vector Magnitude determines how far the transmitted
modulation symbols are from ideal representation.
• 256QAM may require power reduction to achieve a 3,5% EVM.
• The scheduler will decide each TTI dynamically about how much
power backoff is needed to the given TTI
• In comparison to static DL power reduction, the KPI degradation
is smaller for dynamic solution.
• Power back-off aims to reduce the maximum configured output
power with the value that will guarantee the EVM not higher than
3.5% for 256QAM transmission

76 © 2022 Nokia
 LTA18A

DL 256QAM enhancement - Dynamic Power Backoff LTE4054

Field trial result from APJ

before after Expected

KPI Gain
(Mbps) (Mbps) Gain
Avg PDCP cell thp DL 13,34 13.73 2.9% 0-4%

Summary
• Improve average DL Cell Throughput
around 2.9%
• In this trial also show improvement in
the DL User Throughput around 6.2%

77 © 2022 Nokia
 LTE18A
LTE5731
TCP Service Differentiation
Improve the operator network benchmark with Ookla , Opensignal, Rootmatrics, Tutela

• Medium sized TCP data flows are considered to have

highest impact on end user experience.
• Small data transfer can be majorly attributed to
background services thus being less noticeable Medium size
traffic
by end users (e.g. WhatsApp message).
• High data transfers require anyway longer time
for transmission. Avoid rewarding the greedy
users.

• Increased bandwidth allocation for user with medium

traffic size.

• Applied for non-GBR QCI Field test results

• Feature can be activated by the parameter

LNCEL: actTcpServiceDiff. The minimum and maximum Already validated
with more than
number of bytes differentiating medium size traffic is
200 000 cells
configurable.

78 © 2022 Nokia
 LTE18A
LTE5731
TCP Service Differentiation and UL TCP Power Boost Trial (CRL24910) CRL24910

5% gain achieved in OSS stats, 29% based on Ookla speed test

Measurement Period
• before: 20 – 26 Jan 2021
• after: 03 – 09 Feb 2021
Before After
KPI Gain
(kbps) (kbps)

E-UTRAN avg IP sched thp DL, QCI9 5210 5460 4.9%

E-UTRAN avg IP sched thp UL, QCI9 604 634 5.1%

Max Ookla DL test improvement 14614 / 16128 17076 / 18651 16.8% / 15.6%

Max Ookla UL test improvement 8717 / 9165 9174 / 11891 5.2% / 29.7%

Ookla Speedtest Result

The test was conducted on 10 Feb 2021

79 © 2022 Nokia
 LTE18A

Low Latency Fast UL Grants LTE4066

The feature provides two separate functionalities:

1. TCP boost: Non-GBR data triggered proactive UL grants for the UE, so that it can respond to the potential
DL TCP packet without need to wait for Scheduling Request (SR)
2. Transmission skipping on UL grants when Buffer Status Report (BSR) = 0, reducing UL interference and
energy consumption by UE
• Only Rel.14 compliant UEs are able to skip transmission
• Skipping is applicable to all types of UL grants (regular and proactively sent)

Rel.14 UEs can skip TCP protocols can be sped up

transmission of padding data thanks to immediately set ACK
[e.g. TCP Packets]
DL Transmission

UL Transmission
Padding data

Padding data

. . .

[TCP ACK]
UL Grant

UL Grant

UL Grant
[BSR = 0]

[BSR = 0]

TCP ACK preparation

80 © 2022 Nokia
 LTE18A

Low Latency Fast UL Grants LTE4066

Test Reports from NAM

• Speed up TCP procedures and extended functionality for
all types of UL grants allowing them to be “skipped” by
Rel.14 UE Faster browsing
• Improvement to the end user web browsing experience by
enabling UE to respond quicker with TCP positive
acknowledgment (ACK) after getting DL non-GBR data
• Reduced latency and reduction of UL transmission
interference, as well as reduced energy consumption of
UE

Trial results with CRL28236: DL throughput +20%

10% less power consumption

81 © 2022 Nokia
 Load Balancing

82 © 2022 Nokia Confidential

 RL70
LTE2050
Load Triggered Idle Mode Load Balancing
Users Offloaded from loaded frequency Layer to unloaded Layer
Activation
Without LTE2050 With LTE2050 Activated period

RRC Setup Attempt

Frequency layer 1 Frequency layer 2

• Two frequency layers selected. • After feature activated, RRC setup requests, active UEs and PRB
• The load between two frequency is strongly utilization are well distributed to two frequency layers.
uneven: frequency 2 has 22 times attempts • Frequency layer 2 are offloaded significantly. Can take more traffic.
than frequency 1. • Other KPIs in the same level.
83 © 2022 Nokia
 DSS

84 © 2022 Nokia Confidential

 The four flavours of Dynamic Spectrum Sharing
Unique Nokia DSS with interworking

Unique Nokia DSS interworking with RAT specific spectrum efficiency features

GSM-WCDMA DSS GSM-LTE DSS WCDMA-LTE DSS LTE-NR DSS

• BSC Interference Matrix (BIM) used • BIM table used by eNode-B’s to • NodeB dynamically filters DL • Interworking with Nokia Xp
by Node-B’s to dynamically filter dynamical blank LTE in both UL spectrum when no traffic • TDD band dynamic sharing in
out WCDMA in both UL and DL and DL • SBTS sys com interface providing carrier level
• Filtering requires radio HW support • DSS interworking with optional info for eNodeB un-blanking • FDD in-carrier dynamic sharing in
• DSS interworking with optional spectral efficiency features • Uplink is static 3.8MHz PRB level
spectral efficiency features LTE
PRB un-blanking

BIM table BIM table

WCDMA LTE LTE
BCCH BSIC C/I BCCH BSIC C/I SBTS sys com
GSM 514 43 6
GSM 514 43 6 WCDMA
3.8MHz filter
520 61 13 520 61 13 Xp Common resource
BSC BSC 2x0.6MHz
516 23 5 516 23 5 manager in eNB
SBTS 518 1 23 SBTS 518 1 23 SBTS gNB eNB
WCDMA

Increased NW Capacity → CAPEX Efficiency

85 © 2022 Nokia
 Dynamic Spectrum Sharing WCDMA+LTE
Improved spectral Efficiency , Increased LTE capacity , Minimal Re-farming efforts

Network Challenge: Before DSS

• High LTE load, Limited Spectrum

P (dBm/Hz)
• WCDMA is mandatory for Voice service
• RF re-farming challenges.
Solution: 3G 4G
• WCDMA-LTE load shared spectrum: 15 In B1 (2.1Ghz) or f (Hz)
B8 (900MHz) 5 MHz 10 MHz
• Feature : SR002167 (19B) supports: FSMF , 15MHz in B1 15 MHz
• Feature: SR002546 (20A), extend support for Airscale,
B8, BW of 10/15/20MHz
• Maintain WCDMA for voice, expand LTE capacity, while After DSS
maintain WCDMA for voice and legacy U.E
• Configurable threshold for WCDMA Load

P (dBm/Hz)
Benefits:
• CAPEX savings : LTE Capacity Gains ranges from 3G 4G
+12%
f (Hz)
• OPEX Savings : Minimal Re-farming and planning ~3.2 MHz ~11.8 MHz
efforts. 15 MHz

86 © 2022 Nokia
Trial outcome: 3G/4G DSS increased the total payload
Quality maintained across 4G and 3G & 10% Gains in Total DL Capacity

3G+4G gain +9.8% 4G only gain +14% 4G avg connected UE +17.9%

LTE 2100 KPI 3G 2100 KPI

KPI Before After KPI Before After
Total traffic (GB) 3159 3674 +16.3% Total traffic (GB) 485.7 293.6 -39.6%
Cell Availability 95.6% 100% +4.6% Cell Availability 99.8% 99.8% -0.09%
Ave RRC conn. UE 49.2 54.5 +10.75% RAB call block PS 0 0 0%
E-RAB setup SR 99.93 99.94 +0.01 RAB dropped calls CS 0.51 0.23 -54.8%
Block Error Rate Sum of ave user
DL 0.06 0.06 0.00 HSDPA -12.14%
Gain in traffic and quality Decreased traffic, maintained quality
87 © 2022 Nokia
 Dynamic Spectrum Sharing GSM-LTE
Improved spectral Efficiency , Increased LTE capacity , Minimal Re-farming efforts

Network Challenge:
• High LTE load, Limited Spectrum
• GSM is mandatory for Voice service
• RF re-farming challenges.
Solution:
• LTE Slim carrier to support dynamic PRB blanking up to
40%.
• BSC based interference matrix is used to identify
overlapping eNode’s to be notified when shared spectrum is
taken.
• eNodeB blanks corresponding PRB’s and schedules users
only when GSM indicates the freeing up of shared resources
• Configurable Load thresholds
• Reduce Re-Farming & RF Planning efforts
Benefits:
• CAPEX savings : LTE Capacity Gains ranges from +14%
• OPEX Savings : Minimal Re-farming and planning efforts.

88 © 2022 Nokia
Case study: World’s largest commercial DSS network (1/2)
India – 2G/4G DSS

Spectrum Refarming with DSS Introduction

• With >2500 sites deployed and
GSM12.4MHz G444 @ 12.4Mhz rollouts continuing this is likely the
with standard GSM
largest commercial DSS network
today
L900 5MHz G444 @ 7.4Mhz
with DFCA

DSS • VIL: “Nokia’s DSS solution for

optimizing spectrum usage has been
G/L L900 10MHz G/L G444 @ 3.4MHz deployed in multiple circles. This
with DSS+DFCA
1.8MHz (5.4MHz dedicated) 1.8MHz +SBA solution will allow Vodafone Idea to
route the spectrum to different
technology based on traffic
DSS interworking with GSM spectral efficiency features SBA (Smart requirements.”
BCCH Allocation) + DFCA is unique to Nokia’s GL DSS implementation

89 © 2022 Nokia
Case study: World’s largest commercial DSS network
60% Increase in L900 DL Throughput
Average Cell Throughput
Data Traffic (Mbps)
(DL) Mbps

+60% +10%

Pre-DSS Post-DSS Pre-DSS Post-DSS

Instantaneous nature of DSS visible:

• peak shared spectrum triggered in almost all hours
during a day
Daily data volume increased 10% • throughput and payload are increased even during
Quality: DL quality indicators maintained, UL quality improved. GSM busy hours

90 © 2022 Nokia
 Energy efficiency

91 © 2022 Nokia Confidential

RAN energy efficiency features
Up to 20-25 % site level improvement

RF paths micro cuts RF paths switch-off BB capacity switch-off

10% - 25% 10% - 15% 10% - 30% 10% savings

savings savings savings in baseband electricity
consumption
Baseband dynamic
Micro DTX Cell shut down (m)MIMO muting
Data Signaling
power management
High load RRH/RFM mMIMO ABIA card High load
64T64R
RF MIMO4x
power
Time Low load
Low load
32T32R
Enhanced Micro DTX MIMO2x

Data Sleep mode

12% power reduction for
mMIMO when a cell is shut down
Slots
Savings estimated in ETSI average load -- 8 hours busy load + 10 hours medium load + 6 hours low load
92 © 2022 Nokia
 LTE15A

LTE1103: Load based Power Saving for multi-layer networks

LTE1103

Save energy in case of low load periods

Multi-layer eNB monitors DL and UL load, triggers cell Daily

traffic
shutdown once all the following operator adjustable normal load
load
low load
conditions have been met in a cell:
• Time of day
• No emergency calls or ETWS / CMAS in cell, if available
• Traffic load in the cell goes below the defined threshold
• Duration of low traffic conditions longer than threshold
SR002216: Load based Power Saving for
• Aggregated load of the remaining cells in power saving group
does not exceed high load threshold
AirScale SM
• Minimum one ABIA hosting LTE cells

After cell shutdown, the eNB monitors the load on the

CB007416: Cell Switch Off for TDD mMIMO
active cells. A shutdown cell can be resumed when the • CPRI/eCPRISave energy of mMIMO TD-LTE
following conditions are true: solutions during very low load periods
• Traffic load in the active cells exceed the defined threshold
• Duration of high traffic duration longer than threshold

Feature parametrization needed to minimize KPI effect! Expected gain at low load: up to 64% for FDD RRH/RFM
93 © 2022 Nokia
 LTE15A

Field result for Load based Power Saving for multi-layer networks
LTE1103

Up to 9% energy savings with KPIs maintained

Trial setup:
Customer trial:
frequency 3
Power saving checking period: 00:00 AM – 6:00 AM
every day
Turn off condition and sequence for F2 and F3:
• Whole sectors non-GBR load < 3 % → turn off
• Whole sectors non-GBR load > 12 % → turn on

Three sets of cells that cover the same area. Under

low traffic conditions one layer might be switched off
without decreasing network performance.

Feature Off On

Avg energy
2.64 kWh 2.4 kWh
consumption

→ 9% power saving on NW level

94 © 2022 Nokia
 LTE16

LTE1203: Load based Power Saving with Tx path switching off

LTE1203

Save energy during low traffic with MIMO muting

X-pol. antennas
X X
Depending on the cell configuration, one or more radio unit X X
X X
transmitting paths (TXs) can be used to provide DL signal to X X
end users
• The more the transmitting paths, the higher gain in, e.g., cell DL RF Module

throughput is expected (higher cell capacity)

• The more the transmitting paths, the more energy is consumed
to power up each of power amplifiers (PA)

Number of active TX paths is the trade off between

the cell capacity and power consumption

SR003080: Power saving with online TX path

switching off
• The switching is supported online, to avoid end
user impact, MIMO config is kept, no cell outages
due to cell resets (SRAN21B)

95 © 2022 Nokia
 LTE16

Energy saving feature trial results over multiple RATs from MEA
LTE1203
2G/3G
features
Hourly Energy Saving Stats(ENERGY_CONSUMPTION_IN_RF)
250000

200000
Energy Consumed(Kwh*10^5)

150000
27-May

31% 28-May

100000
29-May

30-May

50000 31-May

1-Jun

2-Jun
0
3 AM 4 AM 5 AM 6 AM 7 AM 8 AM 9 AM 10 AM 11 AM 12 PM 1 PM 2 PM 3 PM 4 PM 5 PM 6 PM 7 PM 8 PM 9 PM 10 PM 11 PM
Time

Up to 31% savings observed in non-peak hours (from ~1.5 kWh to ~1 kWh) and 9% overall energy savings in RF after feature
implementation according to Nokia energy meter feature stats.

96 © 2022 Nokia
 VoLTE

97 © 2022 Nokia Confidential

VoLTE readiness
Mandatory KPIs for planning, features enhance the capabilities

Impact to VoLTE
Category KPI • VoLTE is end-to-end
services
functionality
VoLTE service • Service driven network
Coverage • Cell availability availability, quality and
optimization is required
continuity
• Good LTE network is a
• Radio data bearer setup pre-requisite for VoLTE
Accessibility VoLTE call setup rate
successful rate service
• eRAB drop ratio • LTE performance and
Retainability VoLTE call drop rate
• Radio Bearer Success Ratio quality features set
good basis to build on
• Inter-eNB S1 Handover VoLTE call continuity
Mobility • VoLTE specific radio
Success Ratio and drop rate
features enhance voice
• UL/DL retransmission experience
Data quality VoLTE voice quality
• UL latency
VoLTE voice quality
Data throughput • UL/DL cell throughput

98 © 2022 Nokia
VoLTE features
The CS fallback function allows for a service based redirection from LTE to UTRAN or to GSM during the call setup. Both mobile originated and
CSFB to UTRAN or GSM via redirect LTE562
mobile terminated call setup is supported.
Configurable sequence of BCCH carriers within GERAN
LTE2314 Operator configurable sequence of BCCH carriers within GERAN frequency layer list for CSFB based on UE Context Release.
frequency layer list for CSFB based on UE Context Release
Emergency call is handled through CS fallback procedure to GSM or to WCDMA. The emergency CS fallback procedure applies redirection as the
Emergency call handling LTE22
mobility mechanism.
Robust header compression LTE11 Robust header compressions allows for efficient usages of the air interface.
Support of GBR EPS bearer LTE10 The feature allows to introduce high quality managed voice services in LTE.
IMS emergency sessions LTE572 Support of IMS emergency calls.
PS handover based CS fallback from LTE to UTRAN for mobile originated and mobile terminated call setup is supported. The target cell selection
CS fallback to WCDMA LTE736
is based on fast measurement solicitation.
SRVCC to WCDMA LTE872 Service continuity for VoIP
SRVCC trigger enhancement LTE3290 Delayed SRVCC trigger to avoid SRVCC rejection during VoLTE alerting or pre-alerting phase
SRVCC to GSM LTE873 Service continuity for VoIP
SRVCC trigger enhancement LTE3290 Delayed SRVCC trigger to avoid SRVCC rejection during VoLTE alerting or pre-alerting phase
SRVCC to GSM measurement optimization LTE3127 Improved IRAT GSM measurement report time in SRVCC
VoLTE quality triggered SRVCC to GSM LTE3443 eNode B supports SRVCC to GSM triggered upon VoLTE call quality degradation.
TTI bundling LTE907 TTI bundling allows for better uplink performance at the cell border in LTE and reduced PDCCH signaling load.
Service Based Mobility Trigger LTE1127 The feature allows for a service triggered inter-frequency handover.
RLF triggered handover LTE1617 An additional intra LTE handover trigger is introduced.
RLF triggered HO enhancements LTE3105 The feature enhances LTE1617 RLF triggered HO feature with support of RLF Indication in additional scenarios.
Extended VoLTE Talk Time LTE1406 The UE power consumption for VoLTE calls is improved by aligning the DRX on duration time with the scheduling request timing
Dedicated VoLTE inter/target freq list LTE1942 A dedicated VoLTE target frequency list for inter-frequency handovers is introduced.
VoLTE Call Steering Enhancement LTE3723 Voice quality trigger based intra LTE inter frequency mobility
QCI1 establishment starts a configurable timer which keeps the UE in RRC connected mode even if there is no activity (e.g. during ringtone) until
QCI1 Triggered Inactivity Timer LTE2430
timer expiry.
Service based handover thresholds LTE64 Introduction of dedicated VoLTE handover thresholds for coverage based inter-frequency and IRAT mobility.
Service based handover thresholds II CB007079 A dedicated mobility handling is introduced for UEs using specific QCIs.
VoLTE Uplink Coverage Boosting LTE2098 Improved uplink coverage especially for TTI bundling.
VoLTE UL coverage PUSCH extension CB007297 3GPP Rel.14 PUSCH Enhancements covering PUSCH repetition levels and PUSCH modulation restriction.
Emergency Call Based Mobil. Trigger LTE2503 QCI 1 based traffic steering schemes can be limited to emergency calls.
VoLTE Robustness Enhancements LTE3692 This feature adds several enhancements to improve VoLTE KPIs (Audio Gap Ration, Drop Call Rate)
SRVCC Due to Admission Cntrl Reject. LTE2832 SRVCC to WCDMA triggered upon VoLTE call setup rejection by admission control.
Source rate UL grant assign. VoLTE LTE2774 Automatic adaption of VoLTE uplink grants to the source rate.
VoLTE cover.extension codec rate LTE4244 RAN assistance to dynamic bit rate adaptation of audio codecs to improve coverage for voice services

99 © 2022 Nokia
 LTE11: Robust Header Compression
Major reduction of the overhead being transmitted over the air interface
Why
IP overhead from 60 bytes to <6 bytes
• Improved resource Efficiency Cell capacity Gain : ~ 50% reduction of data volume
• Increase Cell capacity
at the air interface

What
• IP/UDP/RTP headers are compressed to
reduce the big overhead.
• IP/UDP/RTP headers size can be either 320
bits (52% of a packet) for IPv4 or even 480
bits (63%) for IPv6
• this means enormous waste of energy for
transmitting headers instead of speech data

10
0 © 2020 Nokia
 LTE16
LTE2098
VoLTE Uplink Coverage Boosting
Multiple enhancements to increase UL coverage

Why
Increases the UL coverage by up to 2.5 dB
• Improved cell edge VoLTE performance. for UEs in a TTI bundling mode, and by up to
• Better UL coverage 1.5 dB for other UEs.
• Improved VoLTE Retainbility

What
• UL VoLTE service improvements via:
• Sensitivity based selection of #PRB and
MCS if UE enters power limitation (PHR<0).
• Reduction of aperiodic CQI reporting.
• Robust Channel Estimation (CE), improved
Noise Power Estimation algorithm
VoLTE voice quality at cell edge is improved,
and call drop rate is decreased.
10
1 © 2020 Nokia
 LTE16
LTE2098
Field result from North America
12.5% Drop Rate Improved by Better Coverage

• ‘LTE2098 VoLTE Uplink Coverage Boosting’

enabled with default value.
• TTI Bundling feature is off and High data
rate (WB-AMR 23.85kbps) or Low Data AMR
(12.65 kbps) codec will be negotiated
during VoLTE call.

10 ©2020 Nokia
2
 LTE18A
LTE2774
Source rate-based UL grant assignment for VoLTE
Up to15% gains VoLTE UL PRB utilization
• Allows the eNB to detect the actual UL bit rate usage (based on
SR and BSR) and adapts the resource allocation to it
• The eNB adapts the UL grant to actual size needed by the UEs in
order to minimize the padding overhead
• E.g. operator assign to high GBR rates for VoLTE (> 70 kbit/s,
although 41 kbit/s would be sufficient). Without LTE2774 this
results in a too high padding (50….100% more PRBs assigned as
needed).
FL18 PT5.6 CRL21668v2 (32)
30000000
CRL21668

Impact, especially in poor radio conditions: 25000000 Up to15% gains VoLTE UL PRB utilization
• Improved system capacity. PUSCH
resources are used in an optimized
20000000

manner 15000000

• Improved system coverage because 10000000

smaller UL grants require less power in

5000000
transmission.
• Improved mean opinion score (MOS) for 0

5/3/2018 13:00
5/3/2018 19:30

5/4/2018 15:00
5/4/2018 21:30

5/5/2018 10:30
5/5/2018 17:00
5/5/2018 23:30

5/6/2018 12:30
5/6/2018 19:00

5/7/2018 14:30
5/7/2018 21:00

5/8/2018 10:00
5/8/2018 16:30
5/8/2018 23:00

5/9/2018 12:00
5/9/2018 18:30
5/5/2018 4:00

5/6/2018 6:00

5/9/2018 5:30

5/10/2018 1:00
5/10/2018 7:30

5/11/2018 3:00
5/11/2018 9:30

5/12/2018 5:00

5/13/2018 0:30
5/13/2018 7:00

5/14/2018 2:30
5/14/2018 9:00

5/15/2018 4:30

5/16/2018 0:00
5/16/2018 6:30
5/3/2018 0:00
5/3/2018 6:30

5/4/2018 2:00
5/4/2018 8:30

5/7/2018 1:30
5/7/2018 8:00

5/8/2018 3:30

5/10/2018 14:00
5/10/2018 20:30

5/11/2018 16:00
5/11/2018 22:30

5/12/2018 11:30
5/12/2018 18:00

5/13/2018 13:30
5/13/2018 20:00

5/14/2018 15:30
5/14/2018 22:00

5/15/2018 11:00
5/15/2018 17:30

5/16/2018 13:00
5/16/2018 19:30
VoLTE calls

10 ©2020 Nokia LCELLR_PRB_USED_UL_VOLTE_NOKLTE(M8011C208_NOKLTE)(No unit)

3
Mobifone Network performance analysis

1 Network performance benchmarking and analysis

2 Recommended feature's

3 ProACTive program introduction

104 © 2021 Nokia public

ProAPTive
Proactive Architecture
Proactive Performance
Proactive TCO/Sustainability

A Partnership for the Best Network in the Industry

105 © 2022 Nokia

Table of Content

• ProAPTive Introduction (for customers)

• Mode of Operation (internal)

• Detail/Example for each stream (adapt with specific customer examples)

• Stream 1: RAN evolution plan
• Stream 2: TCO/S Optimized Solutions
• Stream 3: Golden Cluster
• Stream 4: Software Introduction Framework
• Stream 5: Network Performance Framework

106 © 2022 Nokia Nokia Internal Use

ProAPTive – A Partnership Framework for the Best Network in the Industry

Understanding Your
strategy, objective, NW
Radio Access Network
challenges TCO/S optimized
Network Evolution Evolution
Planning solutions Strategies and
Proposals

Delivery/regular review Work with you on

of solution deployment network architecture
and NW performance and evolution plan

Continuous
Golden Cluster SW Introduction NW Performance Mode of
Implementation Framework Framework Operation
Propose solutions Demonstrate value on
across full Nokia performance & TCO
Portfolio through golden cluster

107 © 2022 Nokia

ProACTive Benefits

Network Performance Framework

RAN Evolution Plan
- Performance analysis and improvements recommendations for Voice Call Setup, Voice
Cadence on Spectrum Evolution, Service & Architecture, Automation & Operation, Speech Quality, Data throughput DL/UL.
Hardware Evolution and Site Solution Strategy.
- Target Open Signal leadership.
“Architecture Roadmap for your network” “End User Experience Monitoring and Pain Points Identification”

Golden Cluster
Demonstrate best network Architecture, Performance
and TCO
“Always having glimpse of the Future”

Software Introduction Framework

Total Cost of Ownership/ Environmental Social and Governance
Ensure equivalent and more features are activated in alignment with the network
Applying leading edge Design, Deployment and Optimisation solutions to maximise
Architecture.
service and minimise interruption during an upgrade.
“Engaging all levers to maximise the network potential”
“Ensuring the best Utility from your Network”

108 © 2022 Nokia

Radio Evolution Plan, Key Areas

A map of the Radio frequencies planned / possible in

Spectrum
the next 5 years, and what the proposed radio access
Evolution
technology plan for each band.

Services & The services, capabilities, features and architectural

Architecture changes expected to be launched in the next 3 years

Automation & A plan for network automation initiatives and cost

Operations optimization in operations and implementation

Hardware HW capability assessment Vs service, architecture and

Evolution automation needs, new hardware introduction, HW life
cycle management.
Pairing of the overall site design with the hardware
Site Solution
evolution, e.g., power, antenna, site types to optimize
Strategy
the end-to-end radio solution.
109 © 2022 Nokia
TCO/Sustainability Optimized Solutions

Benefits

o Minimizing climate
Scalable site Digital Design
impact while providing configurations for Energy
significant OPEX savings and model sites Efficiency
for large scale networks
o Regular engagement for Energy efficient
best possible TCO for site solutions
architecture and site and software
solutions (incl. energy, End to End
and site auxiliaries) Digital
o Digitalized Deployment Deployment incl.
and Network Design with zero touch
reduced manual efforts acceptance

o Access to TCO best

practice repository
updates 1 Part of new proposals
or 2 Improvement proposals

110 © 2022 Nokia

 TCO/S Optimized Solution
End-2-End Digital Deployment

111 © 2022 Nokia

 TCO/S Optimized Solution
Energy Efficiency Feature Activation (focus on 5G )

o 5G Energy Savings features are part of mandatory feature

recommendation to all CTs under proAPTive program
* o 4G Energy savings features are proposed through SW
Sales push. Field results shared with customer. Will be
included under proAPTive in Q2.
o Customer engagement coordinated through CT for trials
and network activations. Reference materials used to
promote SW release and features with customers under
SIF.

Indosat Antina Starhub Normalized Avg.

ID Feature Name Release KT KO SKT KO LGU+ KO KDDI JP SBM JP Optus AU TPG VF NZ Spark NZ dtac TH
ID SG SG activation, APJ

5GC000430 µDTX for Energy Efficiency 5G20B 100% 48% 64% 0% 0% 76% 1% 0% 0% 0% 0% 0% 0% 43%
Cell switch-off (component carrier) for
5GC000431 5G20B 0% 10% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 2%
Energy Efficiency 45%
Embedded Power Meter for Energy
5GC000433 5G21A 28% 49% 30% 0% 0% 78% 88% 3% 0% 100% 0% 0% 100% 37%
Efficiency
5GC002060 Enhanced Common DRX 5G20A 99% 100% 94% 80% 99% 85% 100% 100% 100% 0% 100% 98% 100% 97%
CT level blended activation => 47% 53% 44% 27% 28% 58% 48% 33% 50% 25% 25% 24% 50%
no no no no no no no no
improve improve improve improve improve
change change change change change change change change

112 © 2022 Nokia * Considering addressable BTS, which are technically possible for energy efficiency features activations.
Golden Cluster (GC)

GC establishment &
Benefits updates in line with
o Reference architecture RAN Evolution Plan
of up to 20 sites to
monitor network
performance trends Reference Cluster for
over time
Update RAN Evolution Apply best practices & best possible TCO,
o Experience potential Plan enhancements in GC architecture and network
network performance
performance
improvements in own
network
o Demonstrating
potential for best
architecture,
Review results and Assess and benchmark
performance and TCO
decide on network outcome with expected
o Access to Nokia’s best wide implementation performance
practices repository

113 © 2022 Nokia

 Golden Cluster Performance Dashboard
• 11 APJ Golden Cluster’s onboarded.

• Anonymized Network Quality Index (NQI) used

to compare between Customer Teams.

• Dashboard tracks and outlines reasons for the

degradation or underperformance.

• Two NQI measures are available: Baseline and

Customer Centric.

Record of
degradation or
underperformance

114 © 2022 Nokia

 Software Introduction Framework

Benefits Features to improve the end user experience –

downlink/uplink/ latency/mobility etc.
o Full benefit of the features
available in the software not Benefits => improved ranking in operator
actualized, due to non- benchmarking, improved subscriber loyalty
usage/activation

Deliverables/Outputs
o Analysis of inactivated Features to deliver energy savings and support
features and their
applicability to the customer
sustainability
network config Benefits => lower OPEX, carbon credits
o Leveraging knowledge of
feature gains experience
from across region/global
o Create proof-points for the
feature gain through golden Features to improve operability and serviceability in
cluster live network
Benefits => faster response time to problem
resolution

115 © 2022 Nokia

 SIF Current Progress and Status

o Current feature baseline covers 5G until 21A/21B

o Revised baseline under preparation for features
until 22R1 and also including 4G
o Feature baseline based on assessment of high
value features applicability to customer network
o ProAPTive team working with CT to drive feature
trials for non-active applicable features, and
higher activation for partially active features
o Working with Net Engg team to provide feedback
how to improve the activation status reliability as
shown by MINT through right correlation of
features/configs
116 © 2022 Nokia * Feature was activated 100% in TPG but had to be rolled back due to impact on 3G KPIs. Under investigation (02410332)
Network Performance Framework
Benefits Monitor & Compare Discover Learn
o Understanding Network
Performance vs.
competition and MN RBC
participating Networks.
o Regular performance
reporting
o Investments with clear
performance impact.
o Network “Pain Point”
Detection

Deliverables /Outputs
o Regular OSS NQI
performance reporting.-
forth nightly incl customer o Baseline NQI and customer o Identify network pain points o Customer involvements and
centric KPI’s centric KPI’s used to compare learnings
o Trial network performance
o Automatic anomaly networks
o Detect additional network o Continuous improvement by
detection
o Crowd sourced comparisons opportunities to improve implementing and analyzing
o Monthly Network performance
benchmarking-
o ML to detect performance customer experience
Crowdsourcing base variations o Product, Global and Regional
o Measure applied changes to knowledge sharing and
o Network performance
GC and PLMN support.
playbook
117 © 2022 Nokia
ProAPTive plan update – EXAMPLES ONLY
ProAPTive key
Progress/milestones Next step Challenges & support needed Status
activity

• Plan is ready, workshop with

RAN Evolution Plan • Continue discuss with customers
customers ongoing.

• Done TCO program kick off, regular • Driveless Acceptance not supported
• Negotiation with customers, Site
meetings ongoing by Customer
solution rules defines according to RFQ
TCO Optimized • Draft model configuration agreed • Remote Site Quality not allowed by
requirements
Solutions • MDT demo cluster completed to Customer
• Introduce Automated Site and Cluster
show benefits of OSS+MDT vs Drive • Remote access not available
Acceptance
Tests.

• Fixed CMCC 2.6G SH 19 sites cluster

Golden Cluster • Further features investigation in 22R1,
for GC reference, 22R1 achieved
implementation 22R2 in preparation.
performance target.

• VoNR performance enhancement

• Done feature list and discussion, lab
SW Introduction features (KPI improvement and
and field test, feature list in • Feature list in RFQ
Framework counters) required by CMCC, but we
specification
are late in roadmap

• DL TP increased 6%(550->582) and

UL TP increased 34%(76->102) in
daytime with SW Rel upgrade 21A-
>22R1 • Further features investigation in 22R1,
Network Performance
• Customer Demo 3CC test achieved 22R2 in preparation.
Framework
2.358G (4SSB) in field
(n41/60M+n41/100M+n28/30M;
AEHB+AHPF, 22R1)

118 © 2021 Nokia

ProAPTive Customer Cadence
Automated
Playbooks Dashboards

Month 3: Customer CXO Review

Month 2: TCO + Performance (Lead: SBMU Head, HoS, CT Head & CDM)
• Review of TCO/S initiatives and align with • Performance review: every quarter
customers • Architecture review: every 6 months
• Review network performance KPIs Head of • TCO review: every 6 months
→ Actions to improve network performance Solutions • Joint actions to improve performance,
Account NPO architecture, TCO/S
Manager Manager
Month 1: Architecture + Performance
• Review of the Radio Evolution Plan
→ Align with customer and radio roadmap Technical
• Review network performance KPIs Manager
→ Actions to improve network performance
Implement into MoO Regular Engagement
as normal way of working
Customer
1. Radio Evolution Plan
2. TCO optimized solutions
CX-BG Nokia Internal MoO MN CT Ops 3. Golden cluster implementation
Funnel Drive Details
Review
4. Software introduction framework
5. Network performance framework
119 © 2022 Nokia