2-ACP SSB Coverage Optimization (Based DT) - 202203

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 29

5G Massive MIMO ACP Coverage Optimization

Course Objectives

 Understand the main challenges of 5G Massive MIMO


optimization

 Master the objectives and methods of 5G Massive


MIMO coverage optimization

 Understand 5G Massive MIMO ACP optimization


principles

 Master the 5G Massive MIMO ACP delivery process


and tool usage

Huawei Technologies Co., Ltd


Contents
page
Massive MIMO Optimization Overview

Massive MIMO ACP Coverage Optimization Principles

Massive MIMO ACP Delivery Guide

Typical Massive MIMO ACP Cases

Huawei Technologies Co., Ltd


Advantages of 5G Massive MIMO

Horizontal Beamforming Vertical Beamforming


m
B ea
Radio Front t am
c a s s t B e ea m
T1 a d ca t B
rooardoadcas
WideBeam
Wi BBrB
d oBroadcast Beam
R1 BBrreB aedacast
gNodeB Br oB a r m Beam
oa dcoaadcas
Baseband dc st B t Be
a s ea am
TN tB m
ea
m
RN

PBCH/SS/CSI-RS/PDCCH/PDSCH
All Channel Support BF , Improve Antenna Gain, Extend Coverage Radius Narrow Beam , Decrease Interference , Improve SINR

Square : Horizontal 8 Square+Building : High Building : Vertical


Beam Vertical 2 beam 4 beam

Support 3D BF, adopt to multi Scenario, support beam costumed Increases SU-MIMO/MU-MIMO flow number and double the
capacity
Challenges Faced by 5G Coverage Optimization
Challenge 1: Massive MIMO broadcast beam weights can be adjusted from one-dimensional to multi-dimensional. There are thousands of
weight combinations for a single AAU. How to select the optimal weight in different scenarios?
According to 3GPP specifications, a sub-6 GHz band supports a maximum of eight SSB-beams for polling
digital
transmission. The maximum number of SSB-beams supported is related to the slot assignment and Horizontal Vertical Digital tilt Adjustable
Digital
azimuth Total
timeslot structure. The number of SSB-beams and the direction of each beam used in each coverage Coverage azimuth
Beamwidt Beamwidt adjustmen Steps of adjustable Adjustable
scenario can be flexibly designed. Theoretically, infinite weight combinations are supported. In actual Scenario adjustmen
h h t range Digital Tilt step Steps
t range
engineering applications, Huawei gNodeBs are preconfigured with 17 typical coverage scenarios and number
support remote adjustment of tilts and azimuths, meeting requirements in common wide coverage and Default 105° 6° -2-13 16 0 1 16
building coverage scenarios.
SCENARIO_1 110° 6° -2-13 16 0 1 16
Slot0 Slot1 Slot2 Slot3
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * SCENARIO_2 90° 6° -2-13 16 -10-10 21 336
                                                                                                               
SCENARIO_3 65° 6° -2-13 16 -22-22 45 720
SSBBea SSBBea SSBBea SSBBea SSBBea SSBBea SSBBea SCENARIO_4 45° 6° -2-13 16 -32-32 65 1040
                                   
m0 m1 m2 m3 m4 m5 m6
SCENARIO_5 25° 6° -2-13 16 -42-42 85 1360
                                                                                                               
SCENARIO_6 110° 12° 0-9 10 0 1 10

Beam0 SCENARIO_7 90° 12° 0-9 10 -10-10 21 210

SSB outsourcing SCENARIO_8 65° 12° 0-9 10 -22-22 45 450

Beam1 SCENARIO_9 45° 12° 0-9 10 -32-32 65 650

SCENARIO_10 25° 12° 0-9 10 -42-42 85 850

Beam2 SCENARIO_11 15° 12° 0-9 10 -47-47 95 950

SCENARIO_12 110° 25° 6 1 0 1 1

Beam3 SCENARIO_13 65° 25° 6 1 -22-22 45 45


Beam6
SCENARIO_14 45° 25° 6 1 -32-32 65 65
Beam4
Beam5 SCENARIO_15 25° 25° 6 1 -42-42 85 85

SCENARIO_16 15° 25° 6 1 -47-47 95 95

Measurement scenario: RAN3.0, 64T, slot assignment 4:1, SSB- Grand


6899
based beam densification disabled total
Challenges Faced by 5G Coverage Optimization
The broadcast beam weight adjustment mode is extended from one-dimensional (horizontal beamwidth) of
traditional antennas to multi-dimensional (horizontal beamwidth/vertical beamwidth/tilt angle/azimuth/beam
densification).

vertical
horizontal wave beamwidth
width narrowing broadening

Default,T6,A0 Scenario_3,T6,A0 Scenario_8,T6,A0 Adjusting


the
azimuth

Enabling beam Adjusting the tilt


densification

Scenario_8,Beam Density Scenario_8,T0,A20 Scenario_8,T6,A20


Challenges Faced by 5G Coverage Optimization
Challenge 2: Both SSB and CSI-RS beam coverage optimization must be considered. CSI-RS used for channel measurement does
not support RSRP and SINR reporting.
 The SSB(SS/PBCH Block) functions: SSB: Cell-level polling transmission. A CSI-RS: UE-level; 64T supports 32
maximum of eight sub-6 GHz beams are beams.
Function Channel Description
supported.
Downstream PSS Clock synchronization, frame synchronization, and
synchronization SSS symbol synchronization

Cell Signal Quality SSS


Used for RSRP/RSRQ/SINR measurement, initial
Measurement beam selection, and RRM measurement.
Used by users to obtain necessary information for
Sending a System network access, including the system frame
PBCH
Broadcast Message number (SFN) and the system frame number (SCS)
used by the RMSI.

 CSI-RS functions and classification:


Function CSI-RS Type Description
The SSB and CSI-RS envelopes differ greatly, and the SSB beam direction can be
NZP-CSI-RS
Channel Quality (Non-Zero Power CSI-RS) The UE reports the following information to adjusted. The coverage difference is obvious.
the eNodeB:
Measurement CSI-IM Call Handove Frequent Low Low
CQI, PMI, rank indicator (RI), and layer Question Low RANK
(CM) (CSI-RS Interference indicator (LI) drop r failure handovers MCS rate
Measurement)
The UE reports the following information for
Poor SSB quality √ √ √     √
Beam Management
NZP-CSI-RS beam measurement:
(BM) L1-RSRP, CRI (CSI-RS resource indicator) Poor CSI-RS quality       √ √ √
Time and
frequency offset TRS (Tracking RS)
Used for precise time and frequency offset The SSB reflects the quality of the broadcast channel, affects the initial access and
tracing.
tracing handover performance of 5G UEs, and determines the coverage on the road and the
(TRS)
Used for mobility management measurement.
entire network. CSI-RS reflects the quality of traffic channels, affects CQI reporting, MCS
RRM/RLM
NZP-CSI-RS The UE reports CSI-RSRP, CSI-RSRQ, and CSI- selection, and rank of 5G UEs, and determines the user-perceived throughtput.
Measurement SINR.

Both SSB and CSI-RS have significant impact on user experience of 5G networks. During network planning and optimization, coverage and
interference of SSB and CSI-RS must be both considered.
Challenges Faced by 5G Coverage Optimization
Challenge 3: 5G products have various forms and differ greatly in capabilities. RF adjustment methods are diversified but their functions are different.
Traditional manual RF Adjustment cannot cope with this situation

SSB CSIRS/PDSCH
Number of 64T 32T 8T 4T 2T
TRx • Beam Scenario • Power
• Digital-tilt • M-tilt
• Digital-Azimuth • M-Azimuth
• E-tilt
Product
         
Appearance
Implemen
Optimization Method tation Implementation Effect Limitation
cost
The SSB and traffic channel Limited by the mounting kit or
Mechanical Tilt High
beams are affected. installation condition
The SSB and traffic channel Limited by the mounting kit or
Azimuth High
AAU and RRU+antenna RRU+antenna, RRU+antenna, beams are affected. installation condition
Product Form AAUs
A+P antenna , Easymacro Easymacro book RRU Tradition Some products do not support
al The SSB and traffic channel this function. In co-antenna
Duplex Mode Electrical tilt Medium
method beams are affected. scenarios, it is complex to
TDD TDD TDD TDD/FDD TDD/FDD
modify the electrical tilt.
Coverage The SSB and traffic channel Limited by the maximum
capability High Higher Medium Low Low beams are affected at the same transmit power and maximum
Power Low
time. The SSB power offset is convergence capability of the
supported. product
Coverage 17 types 9 types 4 types 1–2 types 1 type
Scenario Only the horizontal or vertical 32T and lower-capacity
Coverage
Low beamwidth of the SSB beam is products do not support all 17
Scenario
Partially affected. scenarios.
Digital Supported Supported Not supported Not supported
supported The vertical beamwidth cannot
Tilt/Azimuth
broadcast Only the vertical direction of the be adjusted when it reaches the
Digital tilt Low maximum value. 8T and lower-
Not Partially Partially beam SSB beam is affected.
Supported Supported Pattern capacity products do not
Electrical tilt supported supported supported
support this function,
The horizontal beamwidth
Partially Partially cannot be adjusted when it
Mechanical Supported Supported Supported Digital azimuth Low
Only the horizontal direction of
supported supported the SSB beam is affected. reaches the maximum value.
Tilt/Azimuth Not supported by 4T and
lower-capacity products
Challenges Faced by 5G Coverage Optimization
Challenge 4: How to Optimize Road Coverage and Entire Challenge 5: RF optimization is difficult to implement because
Network Coverage? GSM, UMTS, LTE, and NR share the same antenna system.

At the early stage of 5G commercial use, the number of subscribers is small but most Limited by expensive antenna installation platform resources, operators mostly deploy
of them is high-value users. However, RF optimization is mainly based on DT data.
5G by sharing antennas, including LTE and NR sharing AAU and G/U/L/NR sharing
which can only reflect road coverage situation. Only optimize the road coverage may
passive antennas (or A+P). During 5G optimization, the impact on existing RATs
decrease the real 5G user experience. Therefore, Optimization must to consider both
needs to be considered, increasing the complexity.
the road coverage and throughput improvement and non-road performance not
decrease 3.5G/2.6G, 32T/64T

NR 2.6G/3.5G

Op LTE 2. 3.5G,
tim 6G/3.5
G
a lD 32T
co Tc
m om
pr Co-AAU
eh pa
en ris
siv o n
e
op
Op per erc

t im
ex m m
tim ien ial

um 700/
co

NR 3.5G/700/2100
al ce f use

800/
us or r

3.5G, 8T L1800/2100 900/


er

1800/
U900/2100 2100,
2T/4T
700/800/900/1800
/2100M, 2T/4T

Co-passive antenna Co- A+P antenna


Main Functions of 5G ACP Coverage Optimization
SSB coverage optimization based on DT data: The SSB RSRP and SINR reflect the quality of the broadcast channel and affect the initial access and
handover performance of 5G users. SSB coverage optimization can ensure continuous 5G coverage and basic handover performance.

DT-based CSI-RS coverage optimization①: The CSI-RS RSRP and SINR reflect the quality of the traffic channel, affect the CQI/PMI/RI reporting, MCS index
selection, and greatly determine 5G user throughput. CSI-RS coverage optimization can provide better service experience for users.

Handover chain optimization based on DT data: In the cluster single-user peak rate optimization scenario, comprehensively consider factors such as the
distance, coverage, and quality of surrounding cells, create a handover sequence model for serving cells on the road, and perform RF parameter optimization
based on the handover chain modeling result to obtain a better cluster single-user peak rate.

Road optimization also predicts the coverage of the entire network ②: In the 5G network engineering optimization phase, coverage optimization is
performed based on DT data. This method may cause the coverage of other areas to decrease while improving road coverage. ACP uses both DT data and
BT propagation model simulation data to optimize road coverage and improve the coverage of the entire network.

Coverage optimization based on simulation data②: In the optimization phase of a new 5G network, if DT or MR data cannot be collected, BT propagation
model simulation data is used to optimize the coverage of the entire network.

LTE and NR co-antenna coverage optimization: In the LTE and NR co-antenna deployment scenario, besides NR DT data, DT/MR data on the LTE side is
collected to ensure that LTE and NR coverage is both improved during optimization. For details, see 《
5G MCE Service Solution LTE-NR (Co-AAU) ACP Coverage Optimization Technique Guide 》

Function restrictions:

①. This function is a test feature and requires the support of the Mate 30 terminal. In addition, the base station must be upgraded to RAN3.0 and the periodic CSI-RS
measurement must be enabled. If you need to use this function, contact the HQ for support.

②. This function is a test feature. The BT propagation model needs to be deployed on the GENEXCloud platform. If the BT propagation model needs to be used,
contact the HQ to confirm whether the platform supports this function and assist in BT propagation model calibration.
Contents
page
Massive MIMO Optimization Overview

Massive MIMO ACP Coverage Optimization Principles

Massive MIMO ACP Delivery Guide

Typical Massive MIMO ACP Cases

Huawei Technologies Co., Ltd


Massive MIMO ACP Solution Process
Grid-based data
Input Automatic optimization Solution output and result display
evaluation

Target Setting Before After


Coverage evaluation
Engineering
地市 eNodeB 小区CI
parameters
中心载频 物理小区
的信道号 识别码
经度 纬度
电子下 机械下 总下 方位 天线 是否拉
倾角 倾角 倾角 角 挂高 远小区
Coverage Quality Overlapping Overshoot limit Traffic weight
周口 240184 1 38350 283 114.6644 33.5874 6 2 8 120 30 否
信阳
驻马店
郑州
240721
239946
242791
0
0
0
38350
38350
38350
21
87
27
114.1123
114.034
113.4939
32.0924
32.974
34.9569
6
6
9
6
2
0
12
8
9
30
0
50
17 否
15 否
17 是
Interference evaluation
商丘 240385 2 37900 442 115.6809 34.4188 6 5 11 250 30 否
郑州 242791 2 38350 29 113.4939 34.9569 9 0 9 320 17 是
郑州 244674 0 38100 27 113.4956 34.6867 9 3 12 0 48 否

Overlapping coverage
洛阳 254219 1 38350 90 111.5986 34.2956 6 2 8 180 38 否
新乡 249838 1 38350 110 113.941 35.3214 6 2 8 200 45 否
周口 259289 0 38350 84 114.5638 33.7033 6 3 9 0 45 否

evaluation
Iterative optimization algorithm
驻马店 239958 2 37900 98 114.0311 32.9883 6 2 8 300 24 否
洛阳 246426 2 38100 412 112.35 34.6581 6 6 12 240 33 否
商丘 240400 1 37900 91 115.616 34.4381 6 3 9 120 24 否
信阳 240731 1 38350 301 114.1012 32.0952 6 6 12 130 20 是
新乡 249835 0 38350
DT data
295 114.0081 35.32 6 2 8 0 45 否

Overshoot coverage
evaluation

Rasterize
N
Satisfy
Search the Simulation targets
Y
Electronic map desired (Coverage,
result Quality)
?

Antenna file Existing network

Mtilt Azimuth. Etilt


Adjustment
suggestion Beam
Problem Problem Problema Dtilt DAzimuth
grid grid tic area
Problem
Optimizati
Scenario
atic cell
identificati aggregatio aggregati on
filtering Grid Network
on n on Gain Prediction Cell Level
Level Level

Data is the basis. Algorithms is the key Decision-making is the goal


Key Capability 1: Coverage and Interference Modeling & Path Loss
Modeling
Performs grid-based processing on DT data with a precision of 5m, constructs a RSRP matrix model from each NR cell to each grid, and generate a
SSB coverage and interference relationship model based on NR cell information in the grid.
The path loss matrix is constructed on the the cell information, location information, and RSRP information from DT/ frequency scanning/MDT data.
The path loss matrix is a key factor for predicting coverage changes after antenna adjustment. The path loss modeling has two solutions: fixed path
loss and non-fixed path loss.
 In the fixed path loss solution, Keep the same path loss for coverage prediction before and after RF adjustment
 In the non-fixed path loss solution, calculate the multipath propagation information of radio signals before and after RF adjustment based on
Huawei self-developed beam tracing (BT) model to obtain more accurate path loss changes. The non-fixed path loss solution depends on the
GPU resources deployed on the GENEXCloud platform to implement BT propagation model acceleration.

Path Loss Matrix

PL1

Cell 1
PL3

Cell 3

PL2
Road Grid
NonRoad Grid Cell 2
Key Capability 2: Unique Device-Pipe-Chip Synergy CSI-RS Measurement
Capability in the Industry
3GPP Release 15 (38.331) defines CSI-RS RRM (mobility) measurement. Based on the unique device-pipe synergy capability in the industry, Huawei gNodeBs, HiSilicon
Balong5000 chips, Mate 30 terminals, and Probe drive test tools collaborate to implement CSI-RS RRM measurement defined in 3GPP specifications, the 20A/B version
supports CSI-RS RSRP and SINR measurement (including cell-level and beam-level measurement) for the serving cell and the three strongest neighboring cells. Other main
equipment and terminal vendors do not have specific plans. Huawei is more than one year ahead of the industry.
Note that the CSI-RS RRM measurement capability of the base station and terminal side in the current version (20A/B) is not mature. Therefore, the CSI-RS optimization
solution needs to be further matured.
Estimated
Category Problem Description Current Progress
Resolution Time
Temporary solution: The HiSilicon provides a
After periodic CSI-RS RRM temporary non-standard version. Formal Temporary
Base measurement is enabled, event A3 solution: The 22A version works with the solution: 2020.06
Station cannot be reported. (SSB and CSI HiSilicon B5010 version, and the CSI inter- Official solution:
use the same MearObjectId.) frequency measurement is used to solve the 2021Q4
problem.
2T/4T devices are not supported. In 20B, hybrid networking of 32T and 64T is
Base supported. In 21B, beam mapping solutions
8T/32T/64T hybrid networking is 2021Q2
Balong5000 MATE30 BTS5900 Probe Station
not supported. are added to support various networking
modes.
n ly!
eo
CSI-RS RRM measurement supports
The beam mapping solution is added in
最优SSBindex及对应 邻区1 11 11 Base a maximum of three neighboring
n c
version 21B. B5010 supports one primary 2021Q4
Station cells, affecting the accuracy of CSI-
re
的CSI index集合 10 10
9 9
fe
serving cell and seven neighboring cells.

re
8 8 RS interference modeling.
7 7

al
ern requirements of CSI-RS 3I.
6 6
5 5 The RSRP and SINR of CSI-RS 3I The Mate 40 has accepted the RSRP and SINR
4 4 Terminal
cannot be reported.
i ntreporting
2020Q4

or
3 3 7 3 11 15
2 2 6 2 10 14
1
0
1 5
0 4
0 1 2 3 4 5 6 7 8 9 10 11 12 13
1
0
9 13
8 12
0 1 2 3 4 5 6 7 8 9 10 11 12 13 are limited. CSI-RS RRM
F
The B5000 platform specifications
服务小区
邻区2 第1个slot 第2个slot
measurement is supported for a B5010 has been required to support CSI-RS
Terminal In discussion
11 11 maximum of four cells, affecting the RRM measurement of eight cells.
10 10
9 9 accuracy of CSI-RS interference
8 8 modeling.
7 7
分界线:满足CSI-RS测 6 6 The B5000 platform specifications
量触发条件 5 5
are limited. L3 supports only the A requirement has been submitted to B5010
4 4
3 19 23 3 27 31 CSI-RS RRM measurement of the for reporting the measurement of the eight
2 18 22 2 26 30 Terminal In discussion
1 17 21 1 25 29 strongest beam, affecting the strongest beams. The requirement is under
邻区3 discussion.
0 16 20 0 24 28 accuracy of CSI-RS interference
0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 1 2 3 4 5 6 7 8 9 10 11 12 13
第3个slot 第4个slot modeling.
Key Capability 3: Multi-Optimization Target Setting, Automatic Identification of
Problematic Grids, and Problematic Area Aggregation
Currently, four optimization objectives can be set for massive MIMO ACP, Identify problem grids based on the optimization target, including
including SSB RSRP, SSB SINR, SSB Overlapping, and Overshooting weak coverage grids, overlapping coverage grids, and poor SINR grids.
Coverage. Coverage-First optimization and Quality-First optimization are Converges problematic grids into problematic areas based on the
supported. The optimization objective threshold and weight can be types and distribution of problematic grids, and automatically
automatically generated. generates polygon boundaries of problematic areas.

Problem grid Problematic area Problem area


Optimization identification aggregation selection
Target threshold (automatic/manual) Weight
Objectives
The grids whose Based on the The boundary of the problem
SSB RSRP Threshold x area is extracted by
SSB RSRP performance is lower geographical
Default value: CDF10% AlphaShape algorithm, and the
than the set threshold are distribution, aggregate selection is completed based
problem grids problem grids using the on convex hull extraction and
SSB SINR Threshold CFDP clustering least squares ellipse fitting
SSB SINR y
Default value: CDF10% algorithm
Noise point

Core
Point
1. Minimum Signal Level boundary
point
2. RSRP
SSB Overlapping z
3. Cell Number
Default value: 105 dBm, 9 dB, 2

1. Distance Factor Control


2. Minimum Signal Level point

3. Overshooting Grid Ratio Threshold


SSB Overshooting -
4. Overshooting Grid Count Threshold
n
Default value: 1.2-times inter-site distance, CDF10%, C (t )   Pk Bk ,n (t ), t  [0,1].
i 0
15%, 10

When CSI-RS coverage optimization is selected, the optimization objective will the CSI-RS RSRP item.
Key Capability 4: All Scenarios + Full Beam 3D Antenna File Simulation
The 3D antenna file can indicate the antenna gain of the AAU or passive
antenna in each direction (horizontal 0–359°, vertical 90–90°, and step 1°)
of the three-dimensional space. It is an important input for ACP
coverage prediction.
Massive MIMO ACP (2020Q1) has integrated 30+ types of AAU and RRU
antenna files, traversed all beam scenarios, digital tilts, and digital azimuth
combinations, and considered the impact of features such as slot
assignment and beam densification on the number of beams. The number
of antenna files of a single AAU model exceeds 10,000, supports 3D and
2D display of SSB- and CSI-RS-based beam patterns and full-range
optimization of broadcast beam patterns.
Note: To improve the tool running efficiency, the digital azimuth is
optimized in iteration mode with a step of 5°.

+ = *N=
RF single-TRX antenna file TRX Weight File Single-beam antenna file External antenna file
Key Capability 5: Accurate Prediction of Coverage and Interference
The ACP obtains the initial path loss based on DT data. During iterative optimization, the ACP calculates the antenna gain and path loss changes after
RF or pattern parameter adjustment based on the 3D antenna file and BT propagation model, accurately predicts the RSRP of each cell in each grid
after adjustment, then, predicts the optimized SINR and overlapping coverage rate based on big data modeling.

RSRPafter=RSRPbefore+∆antenna gain- ∆pathloss


SINRafter=10*log(RSRPafter/(Interferenceafter+Noise))

Step 3: Change the path loss in BT propagation


Step 1: Calculate the initial path loss. Step 2: Calculate the antenna gain change. model simulation.

Pathloss = TX Power - Feeder loss Gain before Gain after Path loss before Path loss after
+ Antenna Gain – RSRP_before adjustment: 11 dBi adjustment: 17 dBi adjustment: 127 dB adjustment: 125 dB
AAU
s
Variable reflection path loss

Path loss
The path loss of the direct path remains unchanged.
TX power

The antenna
Adjust the
gain increases Adjust the The path loss
digital azimuth
by 6dB at the digital azimuth at the same
clockwise by 20°.
same position. clockwise by 20°. position
decreases by 2
Example: RSRPafter= ( -95dBm ) + ( 6dB ) - ( -2dB ) = -87dBm dB.
Key Capability 6: Primary Serving Cell Distribution Modeling and
Optimization for Optimal Quality and Throughput
In the cluster single-user peak rate optimization scenario, factors such as coverage, quality, throughput, and distance of
surrounding cells are considered. The distribution model of the best primary serving cell in each road section is built, and RF
parameters are optimized based on the modeling result to obtain better road coverage and peak throughput.
Determine the optimal serving cell of each road section and Iterative Optimization Based on Primary
complete the primary serving cell distribution modeling. Serving Cell Distribution Modeling
1. Candidate cell set: 4. Generate the 5. Online review: The
3. Exclude abnormal During iterative optimization, the pattern and RF
Select the cells whose 2. Quality sorting: optimal serving cell primary serving cell
cells. parameters are adjusted to enhance the coverage of
RSRP difference between Sort candidate cells distribution: Consider distribution modeling
Candidate cells with the target serving cell on each road section and reduce
each DT sampling point based on factors candidate cell ranking result can be reviewed
overshoot coverage, the coverage of other cells. In this way, the distribution
and the strongest cell is such as the SINR, and handover factors online, and the target
discontinuous coverage, of serving cells on each road section after optimization
within 6 dB as candidate rate, distance, and and less than 10 sampling to generate the target serving cell of each road
is close to the modeling result, improving the overall
cells. sector direction. points are excluded. serving cell of each section can be manually
coverage quality and throughput of the road.
road section. adjusted.

Cell RSRP SINR Throughput Distance Candidate Quality


Name (dBm) (dB) (Mbps) (m) Cell sorting

A -81 13 540 349 YES 2

B -84 9 515 349 YES 3

C -86 6 470 420 YES 4

D -101 -6 null 420 NO null

E -95 -1 351 327 NO null

F -83 10 633 327 YES 1

G -91 1 413 559 NO null

H -104 -9 null 559 NO null


Example of ACP iteration optimization
SSB CSI-RS
Optimization Object Threshold Weight
SSB RSRP -100dBm 0.75
Tuning Cell Tuning Cell
SSB Overlapping -105dBm,2Cell,6dB 0.25

1 1 8 1 8
8
6 7 2 3
5 6 7 2 3
5 6 7 2 3 1 2 3 4 5
4 5
4 5 4 5 6
1 2 3 4 6 7 8
1 2 3 4 6 7 8 7 8

Overlapping Weak Coverage Overlapping Overlapping

Before 1st round of iteration optimization 2nd round of iteration optimization

Beam Scenario : S2_H90V6 Beam Scenario : S1_H110V6 Beam Scenario : S2_H90V6


D-Azimuth : 0° D-Azimuth : 0° D-Azimuth : -10°
Coverage fulfillment rate : 87.5% Coverage fulfillment rate : 100% Coverage fulfillment rate : 100%
Overlapping fulfillment rate : 75% Overlapping fulfillment rate : 75% Overlapping fulfillment rate : 87.5%
Finess : 87.5*0.75+75*0.25= 84.375 Finess : 100*0.75+75*0.25= 93.75 Finess : 100*0.75+87.5*0.25= 96.875
Gains of DT-based Massive MIMO ACP Optimization
(1) Reduce the proportion of weak SSB coverage and low SINR, 2. Reduce the SSB overlapping coverage to reduce throughput
reduce abnormal events, and improve the 5G camping ratio. drops caused by frequent handovers.
HO Times:11 HO Times:5
ACP improves SSB coverage in the following ways: ACP reduces overlapping coverage by using the Avg THP:437Mbps Avg THP:451Mbps
 Improves SSB coverage of the serving cell. following methods:
 Increase the horizontal or vertical
 Improves SSB coverage of the serving cell.
beamwidth.
 Reduce the SSB coverage of non-serving cells.
 Decrease the downtilt (weak coverage at the
 Reduce the horizontal or vertical beamwidth.
cell edge).
 Increase the tilt.
 Increase the tilt (weak coverage at the cell
 Change the azimuth to deviate the main lobe.
center).
 Decrease the SSB power offset.
 Adjust the azimuth to align the main lobes.
 Increase the SSB power offset.

Optimized Base Beam Optimized Base D- Optimized D- Base SSB Optimized SSB
Base Beam Optimized D- Base SSB Optimized Base SSB Optimized Cell
Beam Base D-tilt Scenario Beam Scenario Azimuth Azimuth RSRP RSRP
Scenario tilt RSRP SSB RSRP SINR SSB SINR
Scenario
Serving H65V6 H90V12 0 0 -91dBm -87dBm
H65V6 H65V12 6 3 -93dBm -88dBm 1 dB 5 dB Neighboring H105V6 H65V6 0 20 -93dBm -99dBm

3) Change the distribution of road serving cells and select the 4. Increase the CSI-RS RSRP, improve the CQI/MCS, reduce
best cell for each road section to improve the Throughput. overshoot coverage, and improve the rank.
Based on the primary serving cell The unique CSI-RS measurement
modeling and iterative and optimization capability
optimization, select better improves the CSI-RS RSRP and
serving cells for low-throughput further improves the CQI and MCS.
roads to ensure proper This feature identifies and resolves
overshoot coverage cells to reduce
distribution of road serving cells
the RANK decrease caused by
and better DT test throughput.
over-long-distance coverage
Optimized Avg DL Optimize Optimized
Base Beam Base D- Optimized Base SSB Optimized Avg Base M- Base Optimize Base CSI-RS Avg DL Throughput
Cell Beam Throughput Cell d M- CSI-RS Avg CQI
Scenario tilt D-tilt RSRP SSB RSRP RANK Azimuth M-tilt d M-tilt RSRP RANK (Mbps)
Scenario (Mbps) Azimuth RSRP
A H65V12 H65V6 6 9 -85dBm -89dBm 2.3 579 A 230 230 14 20 -81dBm -85dBm 11.5 1.7 365
B H65V6 H90V12 8 4 -87dBm -84dBm 3.5 712 B 240 270 11 11 -83dBm -78dBm 12.9 3 615
Gains of DT-based Massive MIMO ACP Coverage Optimization
The Coverage gain specifications:
Based on the planned coverage rate and overlapping coverage rate target of each site, when the DT coverage rate is lower than 90%, the DT weak coverage rate decreases
by 20% after the engineering optimization is started and two rounds of optimization are performed. If the overlapping coverage rate is higher than 20%, the overlapping

coverage rate decreases by 30% after two rounds of optimization.


Throughput gain specifications:

When only SSB coverage optimization is used, the throughput gain is not committed. When both primary serving cell distribution optimization and CSI optimization are
used, the DL Throughput increases by more than 5% after two rounds of optimization.

Definition:
Cluster DT weak coverage rate: For cluster DT data, the proportion of sampling points whose SSB RSRP and SSB SINR meet a certain threshold (for example, SSB RSRP ≥-91 dBm and SSB

SINR ≥ -3) is called cluster DT coverage rate, and the proportion of sampling points whose SSB RSRP and SSB SINR do not meet the threshold is called cluster DT weak coverage rate.

Overlapping coverage rate in cluster DT: If the SSB RSRP difference between a neighboring cell and the primary serving cell is less than or equal to 9 dB, the SSB RSRP of the neighboring

cell is greater than or equal to -105 dBm, and the number of neighboring cells that meet the preceding two conditions reaches a certain threshold (for example, greater than or equal to 2), this

sampling point is called the overlapping coverage sampling point. The ratio of the overlapping coverage sampling points to all DT sampling points is called the cluster DT overlapping coverage

rate.
Note: For a specific cluster, the threshold and target of the weak coverage rate and overlapping coverage rate in the drive test must be determined based on the site planning target. If the
target is set too high, the target cannot be achieved. If the target is set too low, the initial value is too high, affecting the optimization gains. If the DT result reaches the planned weak coverage
rate or overlapping coverage rate, ACP does not ensure coverage optimization gains.

For example, after APC optimization, the DT coverage rate and overlapping coverage rate of a cluster are improved as follows to achieve ACP gains:

Cluster DT weak coverage rate Cluster DT coverage rate Overlapping coverage rate DL MAC Throughput(Mpbs)

Before Before After Before Before


After Optimization After Optimization After Optimization
Optimization
Optimization Optimization Optimization Optimization

16%
14% 553
20% (relatively decreased by 80% 84% 20% 515
(relatively decreased by 30%) (Increased by 7.4%)
20%)
Contents
page
Massive MIMO Optimization Overview

Massive MIMO ACP Coverage Optimization Principles

Massive MIMO ACP Delivery Guide

Typical Massive MIMO ACP Cases

Huawei Technologies Co., Ltd


Massive MIMO ACP Coverage Optimization Delivery Process and Data
Requirements
The accuracy of engineering parameters and integrity of DT data are very important to the ACP
optimization result. During data sorting, onsite survey and configuration check must be performed
to ensure the data accuracy. Otherwise, the gains may not meet expectations or even negative
gains may occur. The accuracy requirements for engineering parameters are as follows:
Longitude and latitude: ±20 m; Antenna height: ±5 m; Azimuth: ±10°; Mechanical Tilt: ±2°

Note: In rooftop site scenarios, if the distance of each sector antenna exceeds 20m, the longitude
and latitude of each antenna need to be collected.

Data Source
Data Requirements
Type

Collect the last updatedNR cell engineering parameters. Ensure that the NR cell information is
complete and accurate. You can obtain the engineering parameter template required for ACP
NR engineering
optimization from the ACP app when creating a task. In addition, you can use the Engineering
parameters Parameter Check app to automatically sort NR engineering parameters required for ACP
optimization.
Collect the latest eNodeB configuration data to check the accuracy of cell configuration
Configuration
parameters in engineering parameters, including power, PCI, frequency, and pattern
File configuration.

Collect the electronic map within one year. The electronic map must be in Planet format and the
precision is 5 m. The electronic map must contain at least the Clutter, Heights, Vector, and
Electronic map Buildings layers. When the electronic map is used together with ray tracing simulation, the
Buildvector layer is also required. The electronic map can be obtained from the carrier's or
Huawei GIS database. The electronic map uploaded by the ACP must be in .zip format.

The ACP tool has been preconfigured with the antenna file for traversing patterns of 5G
Antenna file massive MIMO antennas. You do not need to obtain the antenna file. You can view the massive
MIMO antennas supported by the ACP tool in the antenna file library.
Collect DT data of the optimization area and export the data into a CSV file. During cluster
DT data optimization, the DT must be performed according to the specifications to ensure the
rationality of the test route and vehicle speed.
ACP-based coverage optimization delivery suggestions in massive
MIMO DT scenarios
In the cluster engineering optimization phase, you are advised to perform ACP optimization in two rounds.

Round1: physical RF parameter optimization


If serious weak coverage, overshoot coverage, or overlapping coverage occurs in a cell due to the following reasons, preferentially adjust mechanical antenna
parameters. If necessary, adjust the antenna installation position. Considering the implementation cost and optimization efficiency, the proportion of cells whose
mechanical antenna parameters are adjusted is generally within 15%.

(1) The angle between the antennas of co-site cells is less


than 90°.
2. The main lobe of the antenna is severely blocked
3) The installation position or direction of the antenna is
inconsistent with the planning.
4) The planned mechanical tilt is too large or too small,
resulting in improper coverage.

Note: In 4G/5G co-antenna (or AAU) scenarios (for example, CMCC 2.6G), collect MDT data on the 4G side and DT data on the 5G side for joint RF optimization to
ensure that 4G coverage does not decrease while 5G coverage is improved, for details, see 《
5G MCE Service Solution LTE-NR (Co-AAU) ACP Coverage Optimization Technique Guide 》
.
Round 2: pattern parameter optimization
After the first round of optimization, the second round of
optimization focuses on broadcast beam pattern to optimize the
primary serving cell of each road section and reduce unnecessary
handovers.
Contents
page
Massive MIMO Optimization Overview

Massive MIMO ACP Coverage Optimization Principles

Massive MIMO ACP Delivery Guide

Typical Massive MIMO ACP Cases

Huawei Technologies Co., Ltd


Operator H country C, the coverage rate of CL108 improved by 14 pp after ACP optimization,
and the DL Throughput Improved by 15.3% (67.5 Mbps) - 1/2.
CL108 is a typical urban scenario. 112 NR 2.6 GHz cells have been deployed. Before the
optimization, the DT coverage rate is 61.7% (standard: SSB RSRP ≥ -91 and SSB SINR ≥ -3).
The overall coverage is poor. Two ACP optimization solutions, with different optimization
target thresholds and weights. ACP optimization is performed based on the MM Pattern
parameter. After the task is executed, two optimization solutions and prediction results are
as follows:
Optimization Object Solution 1 Solution 2
SSB RSRP Threshold: Weight -99 dBm: 1 -95 dBm: 3
SSB SINR Threshold: Weight 1.5 dB: 1 3 dB: 1
SSB Overlapping: Weight -105 dBm/2/9: 1 -105 dBm/2/9: 3
SSB Overshooting -99 dBm -95 dBm

ACP provides pattern


adjustment advice for 36 cells
(32.14%), in which 16 cells are
The RSRP prediction gain of solution 2 is relatively high, involved in beam adjustment.
and the adjustment proportion is relatively low. Therefore, solution 2 is selected

Optimized
Objective Unit Base plane
Solution 1 Solution 2
SSB RSRP dBm 87.03 84.41 83.52
SSB SINR dB 6.56 10.21 9.96
SSB Overlapping Grid/Grid 1.74 1.53 1.5.

Tuning Parameter Solution 1 Solution 2


Beam Scenario 21 16
Digital Downtilt 22. 23
Digital Azimuth 9 6
Operator H country C, the coverage rate of CL108 improved by 14 pp after ACP optimization,
and the DL Throughput Improved by 15.3% (67.5 Mbps) - 2/2.
According to geographic statistics (5 m x 5 m), the average SSB-RSRP improve by 2 dB, the SSB-SINR improve by 1.76 dB, the coverage rate (RSRP ≥-91dBm and SINR ≥ -3dB)
iimprove by 14.23pp, and the MAC DL Throughput improve by 15.3% (67.53 Mbit/s). After the ACP optimization, the number of RRC connection reestablishment times decreases
from 5 to 1.

The number of RRC connection Comparison of DL MAC Throughput


reestablishments decreases 4

0921 ACP Before 0921 ACP after


KPIs
Optimization optimization
Serving SSRSRP (dBm) -87.46 -85.47
Serving SSSINR (dB) 6.83 8.59
Coverage of Center (RSRP ≥ -88 and SINR ≥- 3) 55.14 66.57
SINR Prediction Change vs SINR Test Change Coverage of Universal (RSRP ≥ -91 and SINR ≥ -3) 61.7 75.93
NR 1 Stream Rate (%) 21.05 12.1.
NR 2 Stream Rate (%) 52.99 57.64
NR 3 Stream Rate (%) 23.19 - 23.19 26.58
NR 4 Stream Rate (%) 2.77 3.67
NR DL IBLER (%) 9.43 8.91
NR DL MAC THR (Mbit/s) 440.75 508.28
NR DL PRB Number 369626.21 386455.66
NR DL RBLER (%) 0.37 0.15
NR DL Retrans Rate (%) 10.27 9.71
NR Rank Indicator Expects 2.05 2.18
NR Wide Band CQI 12.46 12.82
NR DL MCS Experts 19.06 20.13
Handover Delay (ms) 25.32 26.1
Handover Success Rate (%) 99.4 100
IntraFreq Handover Success Rate (%) 99.4 100
The SINR prediction improvement area is basically IntraFreqHandover Delay (ms) 25.32 26.1
the same as the actual test improvement area. NR Residence Time(s) 3412.76 3470.65
Operator H Country C, SSB-SINR improve by 2.5dB, and the E2E efficiency is 41% higher
than Manual Optimization.
In the high-quality area of Hexi cluster in HangZhou Wulin, 5G massive MIMO ACP According to the ACP optimization result, the SSB SINR improve by
performance gains are mainly verified, and the efficiency comparison between 2.35 dB compared with the baseline, and the manual optimization
manual optimization and 5G ACP is also tested. increases by 2.06 dB.

Before Manual Manual ACP


ACP
KPIs Optimizati optimiza optimization Optimizati
on tion gain optimization
on Gains
Serving SSB RSRP (dBm) 72.27 73.35 1.08 72.12 0.15

Serving SSB SINR (dB) 11.23 13.29 2.06 13.58 2.35


CSI-RS Channel RSRP
76.88 76.25 0.63 76.42 0.46
(dBm)
CSI-RS Channel SINR (dB) 30.58 31.01 0.43 30.67 0.09

NR DL PDCP THR (Mbit/s) 364.12 367.36 3.24 375 10.88

NR DL MAC THR (Mbit/s) 369.03 366.04 2.99 376.07 7.04


ACP performs pattern optimization for 10 cells. The optimized coverage scenarios
Compared with manual optimization and ACP optimization,
include S0/3/4/5/8/14/15, totally 7 types
DT coverage optimization doubles the efficiency of manual
Pattern optimization is manually performed for 14 cells. After optimization, the coverage
scenarios include S0, S2, S3, S4, and S5, total 5 types optimization in terms of data analysis and saves the E2E
ACP
cluster optimization duration by 41%.
Cell Name Baseline Manual
Xinkai Hotel C_65, HZ S0_H105_V6_T3_A0 S4_H45_V6_T9_A15: S5_H45_V6_T5_A15:
Xinkai Hotel C_66, HZ S0_H105_V6_T3_A0 S4_H45_V6_T6_A10:   Three rounds of DT data analysis (manual) – 3 round 750
Xinkai Hotel C_67, HZ S0_H105_V6_T3_A0 S3_H65_V6_T6_A0: S14_H45_V25_T6_A0 5.1 DT KPI Statistics 30
HZ 金海宾馆 C_65 S0_H105_V6_T3_A0 S4_H45_V6_T9_A0 S5_H25_V6_T3_A0
HZ 金海宾馆 C_66 S0_H105_V6_T3_A0 S5_H25_V6_T3_A-10 S8_H65_V12_T5_A0 5.2 Analysis of Abnormal Events (Access Failure, Re-establishment, Call
180
HZ 金海宾馆 C_67 S0_H105_V6_T3_A0 S3_H65_V6_T3_A0 S4_H45_V6_T0_A20 Drop, and Handover Failure)
HZ 器材厂 C_65 S3_H65_V6_T6_A0     5.3 SSB RSRP Weak Coverage Analysis 180
HZ 器材厂 C_66 S0_H105_V6_T3_A0   S0_H105_V6_T6_A0
HZ 器材厂 C_67 S3_H65_V6_T3_A0 S3_H65_V6_T9_A0   5.4 SSB SINR Weak Coverage Analysis 240
HZ light business building C_65 S0_H105_V6_T3_A0 S4_H45_V6_T9_A30: S4_H45_V6_T8_A30: 5.5 Outputting Optimization Advice (Parameters and Patterns) 120
HZ light business building C_66 S0_H105_V6_T3_A0 S5_H25_V6_T3_A20: S15_H25_V25_T6_A0:
HZ Lightweight Business Building C_67 S0_H105_V6_T3_A0 S3_H65_V6_T3_A0:   Three rounds of drive test data analysis (ACP) – 3 round 360
6.1 DT Counter Statistics 30
C_65, No.6 Building, No.1 LY Wulin Building, S0_H105_V6_T3_A0 S2_H90_V6_T4_A0  
Peking University Building, Central, China 6.2 Analysis of Abnormal Events (Access Failure, Re-establishment, Call
180
C_66, Building 4, No. 1, LY Wulin, Building Drop, and Handover Failure)
S0_H105_V6_T3_A0 S5_H25_V6_T9_A20: S3_H65_V6_T6_A0:
Peking University, HZ
6.3 Importing ACP Data (Engineering Parameters, Maps, and Test Logs) 30
No.4 Building C_67, No.1 LY Wulin Building, S0_H105_V6_T3_A0 S5_H25_V6_T6_A30: S5_H45_V6_T9_A0:
Peking University Building, Central, China
6.4 Setting and Execution of ACP Optimization Parameters 60
Suning Electric Appliance C_65, LY Hushu Road,
Peking University Building, Central, China
S0_H105_V6_T3_A0 S5_H25_V6_T9_A20:   6.5 Checking ACP Optimization Advice 60
Thank you. 把数字世界带入每个人、每个家庭、
每个组织,构建万物互联的智能世界。
Bring digital to every person, home, and
organization for a fully connected,
intelligent world.

Copyright©2018 Huawei Technologies Co., Ltd.


All Rights Reserved.

The information in this document may contain predictive


statements including, without limitation, statements regarding
the future financial and operating results, future product
portfolio, new technology, etc. There are a number of factors that
could cause actual results and developments to differ materially
from those expressed or implied in the predictive statements.
Therefore, such information is provided for reference purpose
only and constitutes neither an offer nor an acceptance. Huawei
may change the information at any time without notice.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy