5G Service Experience–Based Network

Planning Criteria

In partnership with
 Contents
Overview................................................................. 4 Cloud VR game experience–based network
planning criteria................................................... 18
In brief................................................................4
Ovum view..........................................................4 Key service features of cloud VR games.........18

Key messages....................................................4 Requirements of cloud VR game experience

on throughput, delay, and packet loss rate.....18
5G service opportunities......................................... 4 Cloud VR game experience–based wireless
network planning criteria................................20
Streaming video, including cloud-based DVR...4
AR and VR...........................................................5 4K live broadcast experience–based network
Online gaming....................................................5 planning criteria................................................... 20
Features of the 4K live broadcast service.......20
The need for network standards............................ 5
Service quality requirements of 4K live
Where to apply standards..................................5 broadcast..........................................................20
Role of standards in network build 4K live broadcast experience–based
and service selection.........................................5 wireless network planning criteria..................23
Benefits of standards.........................................6
Appendix............................................................... 23
Service experience–based 5G network
Further reading................................................23
planning criteria introduction ................................ 7
Overview of service experience–based 5G
network planning criteria..................................7
E2E planning principles based on service
experience..........................................................9

Three-layer EQN mapping modeling from

user experience to network capabilities................ 9
Layer 3: User experience model......................10
Layer 2: Service quality requirements............11
Layer 1: Network capability baseline..............14

VR video experience–based network planning

criteria.................................................................. 16
Main types and service features
of VR videos......................................................16
Requirements of VR video experience
on the throughput and delay............................17
VR video experience–based wireless
network planning criteria................................17

© Copyright Ovum 2019. All rights reserved.

The contents of this product are protected by international copyright laws, database rights and other intellectual property rights. The owner of these rights is Informa Telecoms and Media
Limited, our affiliates or other third party licensors. All product and company names and logos contained within or appearing on this product are the trademarks, service marks or trading
names of their respective owners, including Informa Telecoms and Media Limited. This product may not be copied, reproduced, distributed or transmitted in any form or by any means without
the prior permission of Informa Telecoms and Media Limited. Whilst reasonable efforts have been made to ensure that the information and content of this product was correct as at the date
of first publication, neither Informa Telecoms and Media Limited nor any person engaged or employed by Informa Telecoms and Media Limited accepts any liability for any errors, omissions
or other inaccuracies. Readers should independently verify any facts and figures as no liability can be accepted in this regard - readers assume full responsibility and risk accordingly for their
use of such information and content. Any views and/or opinions expressed in this product by individual authors or contributors are their personal views and/or opinions and do not necessarily
reflect the views and/or opinions of Informa Telecoms and Media Limited.

2 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

About the authors

Daryl Schoolar Paul Lambert

Daryl Schoolar is a practice leader of Ovum's Paul Lambert works in the Europe team at Ovum
Intelligent Networks team, specializing in the mobile as a senior analyst, focusing on the telecoms
access network and, more specifically, technology industry. He specializes in operator strategies, new
and vendor trends around 2G, 3G, 4G (LTE), and 5G. technologies, and mobile broadband best practice.

His research focuses on the radio access market With a keen interest in 4G, Paul continually
(RAN), including traditional macro base stations, evaluates the options open to operators moving to
small cells, and carrier Wi-Fi. In these areas he 4G voice and what this means in the context of OTT
closely follows LTE and 5G developments as well as players' focus on enhancing their core telephony
monitoring activities in 2G and 3G. Daryl's coverage features.
also includes vendors and technology roadmaps.
He specializes in the implications resulting from
Before joining Ovum, he was an analyst at Current eSIM and the evolving capabilities of mobile
Analysis and In-Stat. Daryl has also worked at Cable broadband. Other recent reports have examined
& Wireless in the US as a member of the market eSIM, 5G, LoRA, and IoT.
and competitive intelligence team.

Wang Nanbin Zang Liang

Wang Nanbin works in Huawei Technologies as Zang Liang joined Huawei in 2002 and specializes
Service Fellow, mobile network chief architect, 5G in the field of mobile network planning and
service and experience chief architect, and 5G key optimization. He has rich experience in 3G, 4G, and
projects commercial assurance owner. 5G RAN algorithm research; projects delivery; and
key technical issue solving.
He specializes in the fields of 3G, 4G, and 5G
network technology and the video/cloud VR As the service architect of mobile video scenarios,
experience, key areas in the Huawei GTS global he has accumulated a lot of experience in the design
technical field. and development of video, cloud VR, and cloud game
service experience fields, including experience
He joined Huawei in 1999. As the technical director, evaluation, planning, optimization, and experience
he has successfully led the technical delivery of management.
many corporate-level global projects in the past 20
years. He has accumulated experience in project
delivery and solution architecture design.

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 3

 Overview
In brief
5G encompasses many different technologies, network build options, and services. The decisions
an operator makes when it comes to technologies and build options will greatly impact service
performance. Because of this relationship, mobile operators need a set of network performance
standards to help them select the right technologies and build options to make sure their 5G networks
support the services they want to deliver.

Ovum view
The commercial 5G experience will be an evolutionary one. Mobile operators will start with basic
connectivity services, enhanced mobile broadband (eMBB), and fixed wireless access (FWA). As technology
and markets evolve, 5G will enable those operators to deliver even more advanced and value-added
services. Network planning and deployments must be done in a way that matches the service ambition; this
includes understanding performance requirements and building to network standards that support those
requirements.

The standards process should begin before starting to build the 5G network. Operators should make it part
of their service strategy process and work on enabling as many of those standards as possible on their
existing LTE network before deploying the 5G access network. Network performance standards, however,
are not static. They will evolve during the lifecycle of the 5G network to meet new use cases and customer
performance expectations.

Key messages
• 5G services are in the early stages of development, and service providers need to continue to invest,
collaborate, and partner to ensure they develop the kinds of tailored services that make the most of 5G's
expanded capabilities.
• Mobile operators need to build their network to support different quantifiable performance standards
to ensure they can support different 5G use cases and provide a high quality of experience with those
services.
• Operators should develop a 5G service strategy prior to deploying their 5G network. This way, they can
build the network to the correct performance standards instead of trying to retrofit their 5G later on to
meet service requirements.

5G service opportunities
Streaming video, including cloud-based DVR
Among the main use cases that 5G is set to significantly improve is how customers experience their
networks, especially around video services – not simply as currently experienced but also including
enhanced video experiences that incorporate augmented reality (AR), virtual reality (VR), and the ability to
switch from multiple cameras in real time.

Video performance is the most critical focus area for 5G, mainly because mobile video has grown to
represent 67% of total cellular traffic in 2017 and is set to increase to 83% of total mobile data traffic by
end-2023, according to Ovum research. Unsurprisingly, service providers and their partners are working
toward improving the video experience to take advantage of 5G's speed and capacity benefits over 4G.

4 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

AR and VR
AR and VR are among services that 5G's capabilities will make more attractive to end users because of its
lower latency than 4G as well as its greater capacity and faster speeds. Indeed, this potential is already
being made real. South Korea's LG Uplus is seeing AR and VR services drive usage on 5G networks, with
average data usage of 1.3GB per day compared with 400MB per day on LTE, an increase driven in part by an
increase in the use of AR and VR services.

Online gaming
Online gaming has already seen significant activity among service providers looking to demonstrate the
advantages of 5G over 4G, particularly its lower latency, which online gaming is ideally suited to showcase.
Indeed, some 5G operators, such as Sprint, Verizon, Vodafone, and EE, have launched or plan to launch
online gaming services from leading gaming developers including Hatch and Niantec to both differentiate
in the market around exclusive content and give consumers a clear reason to move to 5G. Partnerships
around 5G gaming will give service providers the opportunity to reinforce their brand in the online gaming
and esports community, for instance around tournament sponsorship, a route Vodafone has already gone
down with its ESL esports partnership. Moreover, moves from major companies such as Google and Apple
into subscription mobile gaming services are set to propel online gaming further into the mainstream and
accelerate 5G uptake.

The need for network standards

Where to apply standards
There are several network areas where operators need to enforce basic levels of performance standards.
Meeting these standards gives operators quantitative network key performance indicators to help them
build a better understanding of service key quality indicators, which help operators improve the overall
quality of experience. Areas where operators need to apply a minimum level of performance are
• latency
• stall
• packet loss
• bandwidth
• minimum speeds.

Of course, performance standards vary by service or application. Streaming video will have different
network performance requirements than a more interactive experience such as online gaming or virtual
reality will. There should be a tight relationship between an operator's network performance and its service
strategy.

Role of standards in network build and service selection

There are two approaches operators can take in offering services using network performance standards. An
operator can start with developing its 5G service strategy, set network requirements based on performance
standards, and then build network to meet those requirements. The other approach is the exact opposite.
The operator builds the network, measures network performance, and then deploys services where the
network meets performance standards.

Ovum advises taking the first approach, because it gives the operator more control and helps to maximize
the network investment. However, network performance and service offering are not locked in with the
second approach. An operator can continue to evolve network performance to support new services. Figure
1 below illustrates this relationship.

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 5

 Figure 1: Relationship between standards, network build, and service selection

Set network Build network to

Select services performance meet performance
requirements requirement

Measure Deploy services that

Build network performance fit performance

Source: Ovum

Benefits of standards
The overall benefit of building a mobile network to a set of performance standards is that it ensures
a high quality of service for the end user. This will keep end users happy and discourage subscriber
churn to competing operators. It will also create a positive reputation for the mobile operator and its 5G
services, which will encourage new end users to subscribe. Ultimately, this will help in monetization of 5G
investments.

Building to quantifiable standards also benefits the operator in a multivendor environment. Commonly,
operators choose more than one radio, backhaul, and core vendor. Making all these vendors meet a
common performance standard will ensure a consistent end-user experience throughout the operator's
footprint. This concept can be extended internationally as well. An operator wanting to offer multinational
service to enterprise customers can partner with other global operators that also build to a common
standard. This will allow for strong performance-based multinational SLAs.

6 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

Note: The following sections of this white paper are supplied by Huawei and may not reflect the views of Ovum.

Service experience–based 5G network planning criteria introduction

Overview of service experience–based 5G network planning criteria
With the evolution of the 5G technology and mobile networks, mobile networks need to be transformed
from the traditional telecoms service–based network construction model to the service experience–
based network construction model (e-planning model) that meets the new digital service experience
requirements.

The core of the e-planning model is the six-order logic, including service feature analysis, experience
modeling and standardization, experience inflection point and baseline, network impact analysis,
experience-based network planning criteria, and experience management (visualized, manageable, and
guaranteed). The e-planning model helps implement modeling based on three-layer mapping, that is, from
user experience model (experience) to service quality requirement (quality) and then to network capability
baseline (network).

Figure 2: Experience-based network planning criteria based on the e-planning model

Service Experience Model Service Quality Requirement Network Capability Baseline

Human factors engineering Linear / Logical regression model Piecewise fitting method

Ultimate
KQI Level Throughput Coverage
Excellent
• Throughput
Good Delay Capacity
• Delay

ITU P.910, P.911, P.913, P.920, ITU-T G.1011

Subjective & Objective Correlation KQI & KPI Correlation KPI & Network Correlation

Nonstalling ratio vs. TBR (throughput-bit rate ratio) Throughput vs. SINR
Initial
Experience No Stalling 100 1,200
Buffering 90
Level Ratio 1,000
DL Throughput (Mbit/s)

80
Nonstalling ratio (%)

Delay 70
90% 95% 98%
800
60
50 600
Ultimate 1s 98% 40
400
30
20
200
Excellent 2s 95% 10
0 0
-7 -1 0 2 10 20 30 33 35
9

0
.

1.

1.

1.

1.

1.

1.

1.

1.

1.

1.

2.

.
<0

>2

Good 3s 90%
9-

0-

1-

2-

3-

4-

5-

6-

7-

8-

9-

CSI SINR
0.

1.

1.

1.

1.

1.

1.

1.

1.

1.

1.

TBR

Source: Huawei

The following network planning criteria are established for typical 5G service scenarios based on
• feature analysis and experience modeling of mainstream 2C services (cloud VR and 4K videos) and 2B
services (4K live broadcast pushing and video surveillance) of 5G
• a large amount of lab and live-network test data
• three-layer EQN mapping model from service experience to network capabilities.

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 7

 Table 1: Overview of network planning criteria for typical 5G service scenarios
Service Network planning criteria
scenario
Scenario Typical Device Throughput Packet loss Millisecond- Delay Coverage Coverage
application requirements rate level peak level CSI quality
instance (typical bit rate) value RSRP CSI SINR
2C 4K on-demand 5G CPE+large 30Mbps 10-3 N/A (without RTT <100ms -113dBm -2dB
video TV screen (VBR: (downlink) affecting
15Mbps) services)
4K live video 5G CPE+large 40Mbps 10-3 (with FEC) N/A (without RTT <100ms -110dBm -1dB
TV screen (VBR: (downlink) 10-5 (without affecting
20Mbps) FEC) services)
4K 360-degree All-in-one VR 80Mbps 10-3 N/A (without RTT <100ms -108dBm 1dB
VR panoramic machine (VBR: (downlink) affecting
video (VR IMAX 40Mbps) services)
and VR concert) Mobile 30Mbps 10-3 N/A (without RTT <100ms -113dBm -2dB
phone+mini-VR/ (downlink) affecting
glasses (VBR: services)
15Mbps)
8K FOV VR video All-in-one VR 100Mbps 10-5 (TCP) N/A (without RTT <25ms -107dBm 2dB
(2D) machine (downlink) 10-4 (UDP) affecting
Mobile services)
phone+mini-VR
(VBR: 50Mbps)
3K cloud VR All-in-one VR 100Mbps 10-5 (TCP) 500–700Mbps RTT <25ms -107dBm 2dB
(Game) machine (CBR: (downlink) 10-4 (UDP)
50Mbps)
2B Surveillance/ HD camera+ 5Mbps 10-3 (with FEC) 16–20× burst RTT <50ms -105dBm 3dB
UAV pushing pushing device (uplink) 10-5 (without throughput
(1080p) (VBR: 2.5Mbps) FEC)
4K live news LiveU pushing 40Mbps 10-2 (with SRT) 16–20× burst RTT <50ms -95dBm 5dB
pushing (30fps) backpack (VBR: (uplink) 10-5 (without throughput
20Mbps) SRT)
4K live Cogent and 63Mbps 10-2 (with SRT) 16–20× burst RTT <50ms -91dBm 8dB
broadcast digital video (uplink) 10-5 (without throughput
pushing at CCTV pushing SRT) About
Spring Festival backpack (CBR: 700–800Mbps
Gala (50fps) 42Mbps)
8K live OSDE (CBR: To be – – – – –
broadcast 120Mbps) constructed
pushing
5G campus – To be – – – – –
constructed
FWA private line – To be – – – – –
constructed
Cloud PC – To be – – – – –
constructed
Telemedicine/ – To be – – – – –
education constructed
Source: Huawei

Remarks:
• Table 1 lists the network planning criteria formulated for the air interface that is the key bottleneck. For
details about the end-to-end (E2E) planning principles, see "E2E planning principles based on service
experience."

8 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

• These network planning criteria are based on the air interface load below 20% (light load scenario). The
standards for medium-load and heavy-load scenarios will be continuously improved after the 5G network
is developed to a certain scale and the live-network data of such scenarios can be obtained.
• In this table, VR 360-degree videos are tested at a fixed point, with a 98% stalling-free rate. Because of
platform problems, 8K FOV VR videos have not been tested.

E2E planning principles based on service experience

Unified planning and domain-based design is the core principle, which implements effective cross-domain
collaboration and domain-based design.

In E2E planning of service experience–based network construction, the requirements of user experience on
the network can be mapped to the baseline requirements such as the E2E throughput, delay, and packet
loss rate. Therefore, the E2E planning of service experience–based network construction uses the unified
core quality of service (QoS) parameters (throughput, delay, and packet loss rate) as the baseline, and the
planning of the wireless network, bearer network, and cloud core network are associated with each other to
implement unified planning and domain-based design of the E2E throughput, delay, and packet loss rate.

Three-layer EQN mapping modeling from user experience to

network capabilities
This chapter describes the modeling methodology from 5G service experience objectives to network
capabilities, including mapping methods and examples.

The three-layer EQN mapping modeling from the user experience model to the network capability baseline
is completed through the two-step hierarchical mapping modeling process from experience to quality
and then to networks. Because of the service experience model and differences between layers, mapping
methods between different layers may be independent from each other.

Figure 3 shows the three-layer EQN mapping modeling from the user-experience model to the network-
capability baseline.

Figure 3: Three-layer EQN mapping modeling

Cloud VR/Video
Experience

Layer 3
Experience Video/Audio Quality Integrity
Interaction Quality
model （Media Quality） (Presentation Quality）
Resolution/bitrate/ MTP/initial buffering Slicing/stalling/
frame rate/FOV latency freezing

Layer 2 TCP/UDP Throughput E2E RTT Packet loss rate

Quality
requirement

Layer 1 Wireless Network Bearer Network Core Network

Network
baseline Spectrum: bandwidth Bandwidth: Bandwidth：
Coverage: RSRP transmission bandwidth CN bandwidth
Interference: SINR Capacity： Capacity：CPU load
Capacity: PRB usage bandwidth utilization

Source: Huawei

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 9

 Layer 3: User experience model
Cloud VR experience
The following three key indicators are used to comprehensively evaluate cloud VR user experience based on
the VR experience evaluation model suggestions provided in ITU-T G.QOE-VR and 3GPP TR 26.929: media
quality index (MQI), interaction quality index (IQI), and presentation quality index (PQI).

• MQI: Media quality score (0–100). This indicates whether the sensory stimulation to users brought by
the VR content, including the audio, video, and content degrees of freedom (DOF), has been close to the
sensory effect in the real world.
Factors that affect media quality include resolution, frame rate, bit rate, FOV, and PPD. These factors are
related to VR content sources and VR devices.
• IQI: Interaction quality score (0–100). This indicates the experience of interaction between a user and VR
content operations when the user uses the VR service. Interaction delay can cause dizziness, nausea, and
isolation from the feeling in the real world.
The DOF is related to the VR game content sources and devices and is irrelevant to networks. The MTP
latency is closely related to networks and needs to be mapped to the lower layer.

Table 2: Key factors affecting the IQI

Factor Impact
MTP MTP latency refers to the response duration of the video and audio after a user performs an action during VR experience.
latency
DOF DOF indicates the mode in which an object can move in space. It is a key factor that helps users create an immersive environment.
Source: Huawei

• PQI: Presentation quality score (0–100). This indicates the continuous and smooth sensory experience of
users when they use the VR service. Poor user experience refers to artifacts and stalling.
Factors that affect the VR presentation quality include stalling, artifacts, and frame skipping. These
factors are related to network capabilities and need to be mapped to the lower layer.

Relationship between the IQI and MTP latency

If the MTP latency is less than or equal to 50ms, the score (96 points) of the IQI is close to the full score. If
the MTP latency is less than or equal to 70ms, the score of the IQI is 88 points. If the MTP latency is less
than or equal to 80ms, the score of the IQI is 85 points, indicating good user experience. If the MTP latency
is greater than 100ms, user experience is poor. It is recommended that the MTP latency be less than 80ms
and good user experience be considered as a basic requirement.

4K video experience
The mobile U-vMOS standard defined by Huawei also applies to 4K videos.

Mobile U-vMOS = f(sQuality, sLoading, sStalling)

where
• sQuality = f(resolution, bit rate, encoding mode, encoding level)
• sLoading = f(initial video buffering delay)
• sStalling = f(video-stalling rate).

The video resolution, bit rate, encoding mode, and encoding level are obtained from the negotiation result
between the video server and the user device. Therefore, they are irrelevant to network capabilities.

The initial video buffering delay and video-stalling rate depend on the matching between video-quality
requirements and network capabilities.

10 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

The mapping from the U-vMOS to sQuality, sLoading, and sStalling and then to the initial video buffering
delay and video-stalling rate is defined by U-vMOS. The mapping can be obtained by visiting the mLAB
website: http://www.mbblab.com:9090/mobilemos.

Layer 2: Service quality requirements

This section describes how to map the initial video-buffering delay/stalling and interactive VR game MTP
latency to the E2E RTT/data transmission throughput.

Video TBR model

"TBR model" refers to the throughput-to-bit rate ratio model. According to lab and live-network data
analysis, the bandwidth required for video playback depends on the video bit rate. In addition, the bandwidth
is the root factor that affects the buffering delay and stalling.

Based on the analysis of a large number of video-play samples (about 700,000) on the live network, it is
found that the network bandwidth for a type of video meets a certain throughput-to-bit rate ratio to ensure
a certain stalling-free rate.

Figure 4: Mapping between the typical throughput-to-bit rate ratios and stalling-free rates

Nonstalling ratio vs. TBR (throughput-bit rate ratio)

100
90
80
Nonstalling ratio (%)

90% 95% 98%

70
60
50
40
30
20
10
0
<0.9 0.9-1.0 1.0-1.1 1.1-1.2 1.2-1.3 1.3-1.4 1.4-1.5 1.5-1.6 1.6-1.7 1.7-1.8 1.8-1.9 1.9-2.0 >2.0
TBR

Source: Huawei

Video services can be considered as throughput-sensitive services.

Compared to bit rate with a wide fluctuation range, the mapping between the throughput-to-bit rate ratios
and the stalling-free rate is more convergent. Based on this analysis, Huawei proposes the TBR model for
throughput-sensitive video services:

NonStalling_Rate = f
( throughput

bit rate )
This model framework can be used to study the mapping between different types of videos and network
bandwidth requirements. Different types of videos have different requirements on the throughput-to-bit rate
ratio. The throughput-to-bit rate ratio required for smooth video playback is determined by the following
factors:
• Fluctuation of the bit rate. The instantaneous bit rate fluctuates around the average bit rate during video
playback. The instantaneous bit rate is high when the video image is rich in details and scenarios are
frequently switched. In the contrary case, the instantaneous bit rate is relatively low. In addition, the jitter
of the network environment must be considered. Therefore, the bandwidth required for smooth video
playback must be higher than the average bit rate of videos.

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 11

 Figure 5: Typical bit rate for video playback
140
130
120
Bit rate (Mbps)

110
100
90
80
70
60
1 10 20 30 40 50 60 70 80 90 100
Time(s)

Source: Huawei

Currently, over-the-top (OTT) and VR videos are encoded based on the VBR, and the bit rate fluctuates
greatly. Some live TV programs and VR games are encoded based on the CBR, and the bit rate fluctuates
slightly.
• Video buffering duration. For on-demand videos, the player has a certain amount of buffer during video
playback. When there is no network download, the buffer means normal playback can continue for a
certain period (generally 10–30s).

Table 3: Typical features and experience results of on-demand and live broadcast videos
On-demand/ Mode Live Perceived Stalling Initial Maximum Content
live broadcast broadcast initial duration ratio buffering time prebuffering fragment
delay buffering time during duration
delay playback
On-demand – – 1.92s 0.03% 4.2s 125s 5s
Live broadcast Normal delay 25s 1.85s 0.46% 4.45s 13.6s 5s
Short delay 9.5s 1.87s 5.50% 2.4s 5.3s 2s
Ultra-short delay 4.4s 1.89s 19.40% 1.89s 2.6s 1s
Source: Huawei

Remarks:
• mLAB obtains the characteristics and experience results of on-demand and live broadcast videos from
the live-network test of the top two carriers in country D.
• The three live-broadcast delay modes are defined by YouTube.

During the playback of an on-demand program, inbound traffic is sufficient, and the client can buffer a
large amount of video data. This way, temporary insufficient traffic caused by network-quality fluctuation
does not affect the playback. During the playback of a live-broadcast program, insufficient inbound traffic
and buffering policies on the client side result in less buffered data and higher stalling than during an
on-demand program. The initial buffering delay can also be mapped based on the ratio of the stable rate to
the bit rate in the initial buffering phase. In addition, the initial buffering delay is related to the network RTT
and can be mapped based on different RTT ranges.

Cloud VR game MTP model

According to the lab analysis, the experience on cloud VR real-time interactive rendering games is mainly
affected by the action response delay (MTP latency). If the MTP latency is too long, users may feel dizzy
and game freezing may occur. MTP latency is an E2E response delay at the application layer, as shown in
Figure 6. (Cyber Cloud indicators are used as examples. They are slightly different from Huawei cloud VR
indicators.)

12 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

Figure 6: MTP latency of a cloud VR game

Terminal Network VR Cloud/Server

Motion
1 NWDelayHMDToServer 2 GametakePosetime
Prediction
Data uplink latency Motion Gaming
Motion Motion
Detecting Sending Receiving Rendering
3

Physical MTP=NWDelayHMDToServer + GameTakePoseTime + RenderingTime +

RenderingTime
EncodingTime + WaitingSendStartTime + NWDelayServerToHMD + Buffering +
DecodingTime

Stream Perceived MTP=Physical MTP – redundancy angle/motion prediction delay (Oculus

Display motion prediction delay 60–80ms, related to terminals)

Stream Stream Stream Gaming CG

Decoding Buffering Data downlink delay Sending Encoding

8 DecodingTime 7 Buffering 6 NwdelayServerToHMD 5 WaitingSendStartTime 4 EncodingTime

E2E MTP includes：

Related to RTT Related to BW
• Cloud Delay (Server): 2 3 4 5
1 UL TX Delay 5 Server waiting time
• Terminal Delay (HMD): 7 8
6 DL TX Delay
• Network Delay (RTT): 1 6

Source: Huawei

As shown in Figure 6, the E2E MTP model covers the processing delay of three parts, that is, cloud, pipe,
and device:
• Cloud processing delay: includes the VR cloud action obtaining delay (2), rendering delay (3), encoding
delay (4), and waiting for sending delay (5).
• Pipe processing delay: includes the uplink transmission delay (1) and downlink transmission delay (6). The
sum of the two delays is close to the network RTT.
• Device processing delay: includes the buffering delay (7) and device decoding delay (8).

According to the live-network test data, the average value of the total cloud and device-processing delay ((2)
+ (3) + (4) + (5) + (8)) is about 41ms. The total delay can be considered as a constant A.

Constant A = action obtaining delay + rendering delay + encoding delay +

waiting for sending delay + decoding delay

MTP latency = constant A + uplink transmission delay +downlink transmission delay +

buffering delay = constant A + RTT + t

According to the preceding analysis, the following three delays in the MTP latency are directly affected by
the pipe capability: uplink transmission delay, downlink transmission delay, and buffering delay. The uplink
transmission delay and downlink transmission delay are close to the network RTT, but the buffering delay t
(stream buffering) is inversely proportional to the data transmission throughput, as shown in Figure 7.

In Figure 7, t is calculated using the following formula:

t = data amount of a frame/data transmission throughput
where
• t is the data-receiving delay, that is, buffering delay
• data amount of a frame is the amount of data that needs to be sent after the server encodes each game
image frame.

Therefore, in a network, after the server and the device are determined:

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 13

 MTP latency = constant A (about 41ms for cloud-based devices) + RTT +
(game bit rate/frame rate)/network throughput

According to the preceding formula, the MTP latency is directly restricted by the RTT and network
bandwidth after the game bit rate and frame rate are fixed. Therefore, optimizing the RTT and throughput is
the key to reducing the MTP latency.

Figure 7: Buffering delay t and data transmission throughput

600,000

500,000
t
Packets (Bits)

400,000

300,000

200,000

100,000

0
94.5 94.52 94.54 94.56 94.58 94.6

Source: Huawei

Layer 1: Network capability baseline

The sections "Layer 3: User experience model" and "Layer 2: Service quality requirements" analyze the
requirements of service experience on the throughput and delay. However, the requirements cannot be
directly used for radio network planning and must be mapped to coverage and load indicators and used as
the direct input for radio network planning.

Currently, most 5G networks mainly analyze the impact of coverage (level and quality) on the throughput
and delay.

Throughput requirements on 5G network coverage and interference

• Analysis of the relationship between the downlink throughput and coverage/interference. Analyze
the downlink throughput scheduling principle. The downlink throughput is directly related to the SINR
reported by user equipment (UE). Therefore, the downlink CSI SINR is used as a key indicator for
evaluating coverage and interference.

Figure 8 shows the mapping threshold in typical scenarios (3.5GHz, 64T6R, densely populated urban areas,
50% loading load, and 100MHz bandwidth). The mapping threshold is obtained by analyzing the relationship
between the downlink edge throughput and SINR based on the actual test data.

• Analysis of the relationship between the uplink throughput and coverage/interference. The uplink
channel quality of a UE can be indicated by the SINR.

The uplink CSI SINR cannot be tested on the UE side and needs to be replaced by other testable indicators.
According to the theoretical calculation and actual test data, the mapping between the uplink CSI SINR and
the downlink CSI RSRP is available.

Uplink RSRP = downlink RSRP + UE TxPower - gNodeB TxPower.

14 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

That is:
Uplink SINR = downlink RSRP + (UE TxPower - gNodeB TxPower) - (noise + uplink IoT)

Figure 8: Relationship between the downlink throughput and SINR in typical scenarios

1,200

1,000
DL throughput (Mbps)

800

600

400

200

0
-7 -1 0 2 10 20 30 33 35
CSI SINR

Source: Huawei

Requirements of E2E delay on 5G networks

Based on the 5G network architecture, the E2E delay can be divided into four segments: radio network
delay, bearer network delay, core network delay, and server/CDN delay.

Figure 9: E2E delay segmentation on the 5G network

CDN

Cloud VR

Access Aggregation Core Datacenter

EPC/NC

Wireless Bearer Network Core Network CDN

Source: Huawei

In normal cases, the processing delay of the core network and CDN is very short. The delay planning mainly
involves the bearer network delay and radio air interface delay.

Analysis of the bearer network delay

Figure 10 shows the position where the bearer network delay occurs.

As shown in Figure 10, the bearer network delay consists of the transmission delay, propagation delay,
processing delay, and queuing delay. The characteristics of factors causing delays are as follows:
• Transmission delay (plannable). The delay of transmitting data bit by bit through the transmission
medium depends on the size of the data packet to be transmitted and the transmission throughput (bps).
• Propagation delay (plannable). The transmission delay of an optical fiber depends on the refractive index
and length of the optical fiber, that is, the transmission distance.
• Processing delay (measurable and simulatable). The processing delay indicates the delay for processing
data packets inside network devices. For a typical router, the processing delay of each hop is about 20μs
to 30μs when no congestion occurs.

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 15

 Figure 10: Where the bearer network delay occurs
E2E RTT

Bearer Network Delay

Node equipment Transmission path

Client Server

Process delay, TX delay Path (propagation) delay

Planning

✓ OTN: 30μs/set ✓ Fiber propagation delay: 5μs/km

✓ IP: 10~15μs/set
✓ 5G MW: 50μs/hop
✓ TX delay: Packet size/port transmission speed

Queuing delay
Design

✓ QoS marking ✓ Scheduling

✓ Queue / shaping

Source: Huawei

• Queuing delay (measurable and simulatable). The delay in a network device queue usually depends on
the load and congestion degree of a network link, QoS priority of a data packet, and scheduling algorithm.
• Bearer network medium selection suggestions. It is recommended that the 5G bearer network use the all-
optical networking mode. If microwave link backhaul must be used in some scenarios, it is recommended
that microwave link backhaul be used only for accessing a single gNodeB. It is not recommended that
microwave link backhaul be used on aggregation or backbone networks. The E-band 80GHz high-frequency
solution is recommended (bandwidth: 10–20GHz; transmission distance: 1–3km/hop).

Requirements of air interface delay (air interface RTT) on wireless network coverage and capacity
According to the analysis of the relationship between the air interface RTT and the CSI RSCP/SINR, the
impact on the RTT is not obvious if the RSRP is greater than -110dBm; the RTT will deteriorate if the SINR is
smaller than 1dB.

The air interface delay is positively related to the network capacity. Currently, the 5G network is unloaded.
Therefore, the heavy-traffic lab test data needs to be provided to obtain the mapping between the network
load and air interface delay.

VR video experience–based network planning criteria

Main types and service features of VR videos
Cloud VR service scenarios are classified into cloud VR videos and cloud VR games based on the interaction
degree.

Based on the transmission mode and content, on-demand VR panoramic videos use the full-view or FOV
transmission solution. In the current VR panoramic videos, 4K videos use the full-view transmission
solution to reduce the bit rate. The full-view transmission solution is widely used in the initial phase of
cloud VR video services. The cloud pushes all 180- and 360-degree video content to devices, and devices
are responsible for tracing the change of the user head posture and parsing, rendering, and displaying the
locally cached audio and video data in real time. The process from the header posture change to the display
exists only on the device side. This ensures that the MTP latency is shorter than 20ms. Other processes are
the same as those of common OTT videos. Stalling caused by insufficient bandwidth is the main problem
facing this solution.

16 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

Requirements of VR video experience on the throughput and delay
Requirements of 4K 360-degree videos and VR IMAX movies on the throughput and delay
On the 5G commercial network, a series of subjective experience tests are performed for encoding policies
with different content according to the subjective evaluation method in the VR experience model.

Currently, 5G VR videos are at the initial stage. The recommended bit rates are as shown in Table 4.

Table 4: Recommended bit rates for 5G VR videos in typical scenarios

Scenario Resolution Entry-level bit rate Frame rate Encoding format
VR IMAX video 1080p 8Mbps 30fps H.264
4K 25Mbps 30fps H.264
360-degree video 4K 40Mbps 30fps H.264
8K FOV 15+35Mbps 30fps H.265
Source: Huawei

Remarks:
• The data is obtained from the test results of the 5G live network of Zhejiang Mobile and Huawei
convergent video platform. The H.264 encoding mode is used.

Based on the recommended bit rates and live-network test results, the required VR throughput-to-bit
rate ratio is twice the bit rate (stalling-free rate in the static environment: 98%). Based on this, obtain the
required throughput. For details, see "VR video experience–based wireless network planning criteria"
below.

VR video experience–based wireless network planning criteria

VR video experience–based wireless network planning criteria (HLS)
Both 4K 360-degree VR panoramic videos and VR IMAX use the HLS protocol. During playback, service
features of such videos are the same on the network side and are the same as those of 4G on-demand
videos. Therefore, sample data of these videos is combined and processed together.

Wireless network planning criteria based on lab test data (static and noninterference environment test)
According to the VOD test samples obtained from the live network, when the download throughput is 1.4,
1.6, 1.8, and 2.0 times the average video bit rate, the corresponding stalling-free rate can reach 80%, 90%,
95%, and 98% respectively, at least.

Figure 11 shows the fitted curve of the VR video-stalling rate and the throughput-to-bit rate ratio.

Figure 11: Fitted curve of the VR video-stalling rate and the throughput-to-bit rate ratio in the lab environment

VR stalling rate vs. TBR

1.4TBR@80% nonstalling
80
1.6TBR@90% nonstalling
70
Nonstalling rate (%)

60 1.8TBR@95% nonstalling
50
40 2TBR@98% nonstalling
30
20
10
0
0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4

Source: Huawei

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 17

 Remarks:
• The data is obtained from the indoor distributed lab network, NSA networking, and indoor CQT. The signal
is stable. The rate limitation test is performed through the router.
• A total of more than 3,000 samples were collected from April to May in 2019.

According to the previous test results, the recommended bit rate for 4K 360-degree VR videos is 40Mbps.
Table 5 lists the recommended bandwidth and coverage requirements.

Table 5: Specifications for network construction based on VR video experience in typical scenarios in the lab environment
Category Bit rate Stalling-free rate Throughput-to- Bandwidth Coverage quality Coverage level
bit rate ratio CSI SINR RSRP
4K VR 360 video 40Mbps 90% 1.4 56Mbps -0.3dB -110
4K VR 360 video 40Mbps 90% 1.6 64Mbps 0.2dB -109.5
4K VR 360 video 40Mbps 95% 1.8 72Mbps 0.6dB -109
4K VR 360 video 40Mbps 98% 2 80Mbps 1dB -108
Source: Huawei

Remarks:
• For details about the mapping from the bandwidth to SINR, see "Throughput requirements on 5G network
coverage and interference."

Cloud VR game experience–based network planning criteria

With the development of a 5G network and industry, cloud VR games are considered to be one of the
potential mainstream services in the 5G era. By studying the service features and experience of cloud VR
games, we can find the service quality demands (throughput and delay).

There is no sufficient commercial environment for cloud VR games. The sample data used for developing
the current network planning criteria is obtained from the OpenLab test environment. After cloud VR games
are put into commercial use and popularized on the 5G network, the same modeling mechanism can be
used to optimize the experience-based network planning criteria based on the commercial network data.

Key service features of cloud VR games

Cloud VR games are typical cloud-based services. Compared with traditional video services, cloud VR
games have higher requirements on the network bandwidth, delay, and packet loss rate. Cloud VR games
have the following key features:
• Millisecond-level pulse. When the encoding rate is 50Mbps, the millisecond-level peak rate can jump to
200Mbps. Calculate the average number of transmitted packets based on samples collected at the same
time point. Two or even 17 packets can be sent per millisecond. The number differs greatly within the
same time segment.
• Fixed packet length. Game packets are in TCP format, and the length of each packet is fixed at 1,424 bytes.
• Fixed number of transmitted packets. There are 4,800 packets sent per second.
• Fixed encoding rate and frame number. For example, the encoding rate is 50Mbps, 60 frames are sent
per second, and the size of each frame is 0.83Mbps (50Mbps/60).

Requirements of cloud VR game experience on throughput, delay, and packet

loss rate
To verify the impact of the throughput and delay on user experience of cloud VR games, two types of tests
are performed in the OpenLab: one with limited bandwidth and one with unlimited bandwidth:
• The test with limited bandwidth aims to verify the changes of subjective game experience and MTP latency
under different bandwidth configurations and find the inflection point of the impact of the throughput on
VR game experience.

18 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

• The test with unlimited bandwidth aims to verify the changes of subjective game experience and MTP
latency under different packet loss rates and network delays and find the inflection point of the impact of
the packet loss and delay on VR game experience.

Figure 12: Network capability vs. MTP and experience

Experience score MTP (ms)

Network capability vs. MTP and experience (TCP)

7 3,500

6 3,000

5 2,500

4 2,000

3 1,500

2 1,000

1 500

0 0
P s+

P +

PL s+

PL +

PL +

PL +

PL s+

PL +

PL s+

PL +

PL +

PL +

PL s+

PL s+
2X 1% ms

% ms

% ms

% ms

% ms

% ms

% ms

% ms
2X 1% 0m

1% 5m

1% 5m

1% 0m

1% 0m

1% 5m
tra LR

tra LR

R
0 0

01 5

01 5

01 5

01 5

01 0

01 5

01 0
00 2

0. +2

00 2

0. e+2

0. +2

0. +2

00 2

0. e+2

00 2

0. +2

0. +2

0. +2

00 2

00 2
0. te+

0. e+

0. te+

0. te+

0. rate

0. te+
te

te

te

te

te

te
t

t
tra

tra

tra

tra

tra

tra

tra

tra

tra

tra

tra
t
bi
bi

bi

bi

bi

bi

bi

bi

bi

bi

bi

bi

bi

bi
2X
2X

3X

8X

5X

5X

2X

5X

8X

8X

5X

2X
1.
2.

2.

1.

1.

1.

1.

1.

1.
Network capability vs. MTP and experience (UDP)

5.0 100
4.5 90
4.0 80
3.5 70
3.0 60
2.5 50
2.0 40
1.5 30
1.0 20
0.5 10
0.0 0
P s+

P s+

P s+

P s+

P s+

P s+

P s+

P s+

P s+

P s+

P s+

P s+

P s+

P s+

PL s+
2X 1% 0m

2X 1% 5m

2X 5% 5m

5X .1% 0m

2X .1% 0m

2X .1% 5m

2X .1% 0m

2X .5% 5m

2X .1% 5m

5X .1% 5m

2X .1% 5m

5X 1% 0m

2X 1% 0m

2X 1% 0m

1% 5m
tra LR

tra LR

tra LR

tra LR

tra LR

tra LR

tra LR

tra LR

tra LR

tra LR

tra LR

tra LR

tra LR

tra LR

R
0. e+2

0. +2

0. +2

0 +2

0 +2

0 +2

0 +2

0 +2

0 +2

0 +2

0 +2

0. +3

0. +3

0. e+3

+2
te

te

te

te

te

te

te

te

te

te

te

te

te
t

t
tra
0

0
bi

bi

bi

bi

bi

bi

bi

bi

bi

bi

bi

bi

bi

bi

bi
2X

1.

1.

1.

1.

1.

1.

1.

Source: Huawei

The key conclusions of the cloud VR game test on service quality requirements are shown in Table 6.
Table 6: Service quality requirements on cloud VR games
Content Bandwidth Packet loss rate Network delay
Cyber Cloud (TCP) ≥2×bit rate ≤0.001% <25ms
Cyber Cloud (UDP) ≥2×bit rate ≤0.01% <25ms
Source: Huawei

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 19

 Cloud VR game experience–based wireless network planning criteria
Table 7: Key specifications for wireless network construction based on cloud VR game experience
Service scenarios Network requirements Coverage standard

Service type Typical application Device Guaranteed Delay Coverage level Coverage quality
instance requirements bandwidth CSI RSRP CSI SINR
Downlink 3K cloud VR game Cloud VR games All-in-one VR 100Mbps RTT -107dBm 2dB
service bit rate: 50Mbps/ and VR education machine (downlink) <25ms
TCP
Source: Huawei

Remarks:
• The preceding wireless network planning criteria are developed based on data in the OpenLab test
environment. The commercial environment will be continuously improved after services are put into
commercial use.

4K live broadcast experience–based network planning criteria

Features of the 4K live broadcast service
Resolution and bit rate
The 4K standard is also called "ultra-high definition" or "ultra-HD" and is used for displays that have an
aspect ratio of 16:9 or wider and can present native videos at a minimum resolution of 3,840×2,160 pixels.

The bit rate of 4K VOD videos is about 15Mbbps. TV stations and carriers have higher requirements on 4K
video service quality during live broadcast. The average bit rate (about 20Mbps) of 4K live broadcast videos
is about 20–50% higher than that of internet videos with the same resolution and frame rate. CCTV 4K live
broadcast (for example, CCTV Spring Festival Gala live stream) has higher pushing requirements. Generally,
the bit rate must be 42Mbps. Table 8 lists the service features in typical 4K live broadcast scenarios.

Live broadcast timeliness (delay)

Live broadcast timeliness (delay) refers to the delay from video image generation to video consumption.
No unified standard is available for live broadcast timeliness in the industry. Currently, the commonly used
video timeliness rating standards are classified as follows:
• Pseudo real-time (noninteractive) live broadcast. The video consumption is delayed for more than 3s.
• Quasi real-time (weak interactive) live broadcast. The video consumption is delayed for 1–3s.
• Real-time (interactive) live broadcast. The video consumption is delayed for less than 1s, and the average
delay is 500ms.

Currently, OTT noninteractive live broadcast is pseudo live broadcast with long delay.

Broadcast and television stations and carriers have higher requirements for noninteractive live broadcast.
The delay defined for live broadcast is less than 2s. The delay requirements for CCTV 4K live broadcast are
stricter at 1s.

Service quality requirements of 4K live broadcast

4K live broadcast playback (stream pulling)
The 4K live broadcast service uses the UDP transmission mode to meet real-time requirements. If
no forward error correction (FEC) redundancy or retransmission (RET) mechanism is available, user
experience is sensitive to network fluctuation. When the packet loss rate is greater than 10-5, artifacts occur.

20 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

Figure 13: Relationship between the artifact ratio and packet loss rate according to the EDN lab test

Test 1 Test 2 Test 3 Test 4 Test 5 Test 6 Test 7 Test 8 Test 9 Test 10

100
90
80
Artifact ratio (%)

70
60
50
40
30
20
10
0
0 1.00E-06 1.00E-05 1.00E-04 1.00E-03

Packet loss rate

Source: Huawei

In addition, the test result shows that the bandwidth required for watching 4K live broadcast programs is
1.5–2× the bit rate. When the network bandwidth is two or more times the bit rate, no artifact or stalling
occurs.

4K live broadcast (stream pushing)

Because of real-time requirements, 4K live broadcast pushing also uses the UDP transmission mode. If
no reliability protection mechanism is available, 4K live broadcast pushing is also sensitive to network
fluctuation and requires that the packet loss rate be lower than 10-5.

After 2Q19, mainstream 4K pushing backpack vendors, such as Cogent, Sumavision, and LiveU, developed
reliable UDP transmission technologies, including FEC and SRT (ARQ):
• Network bandwidth. When the bit rate is 42Mbps (CBR), if the uplink bandwidth is lower than 50Mbps,
frame loss occurs. If the uplink bandwidth is higher than 50Mbps, frame loss does not occur. Considering
the fluctuation of the uplink throughput of the wireless network, it is recommended that the bandwidth be
1.5× the bit rate to ensure the smooth 4K live broadcast pushing experience.
When the adaptive dynamic bit rate is used, the initial bit rate is 20Mbps. The uplink bandwidth must be
higher than 40Mbps (about twice the bit rate).

Figure 14: 4K live broadcast pushing requirement for PLR

Cogent Sumavision

4
Frame loss rate (%)

0
45 46 47 48 49 50 55
Bandwidth (Mbps)

Source: Huawei

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 21

 • When the uplink bandwidth is higher than 40Mbps, the 4K live broadcast pushing quality is stable.
When the uplink bandwidth is lower than 40Mbps, the live broadcast bit rate decreases adaptively and
the definition decreases. However, the pushing quality is still stable and no artifact occurs. When the
bandwidth is lower than 10Mbps, mosaics occur.
• Network RTT. According to the test results of Cogent and Sumavision, the network RTT required by 4K
live broadcast pushing is about 110ms when the live broadcast delay is set to 1s. If the live broadcast
delay is 0.6s or 0.8s, the RTT is smaller but cannot be smaller than 50ms.
However, LiveU devices have a higher requirement on the network RTT, that is, 50ms when the delay is
1s.
• Packet loss rate. After using reliable UDP transmission technologies such as SRT, Cogent and Sumavision
reduce the requirement on the packet loss rate. That is, before such technologies are used, the required
packet loss rate must be lower than 10-5. After such technologies are used, the required packet loss rate
need only be lower than 10-2. According to the lab test results, the acceptable packet loss rates are 13%,
10%, and 5% respectively when the delay is set to 1s, 0.8s, and 0.6s.
LiveU uses redundancy coding and has low requirements on the packet loss rate. That is, the packet loss
rate only needs to be lower than 8.5%.

Figure 15: 4K live broadcast pushing requirement for PLR

1s delay 0.8s delay 0.6s delay

7
6
Frame loss rate (%)

5
4
3
2
1
0
5 6 7 8 9 10 11 12 13 14 15 16 17 18
Packet loss rate (%)

Source: Huawei

22 Informa Tech Ovum © 2019 Informa PLC. All rights reserved.

4K live broadcast experience–based wireless network planning criteria
Table 8: 4K live broadcast stream pulling requires sufficient downlink bandwidth, and 4K live broadcast pushing requires
sufficient uplink bandwidth
Service Device Typical bit Frame Network Packet loss Millisecond- Delay Coverage Coverage
scenario vendor rate rate bandwidth rate level peak level CSI quality CSI
value RSRP SINR
4K live Huawei 20Mbps 30fps 40Mbps, 10-3 (with FEC N/A (without <100ms -110dBm -1dB
broadcast hybrid video which is 2× coding) affecting
video platform the bit rate 10-5 (without services)
(stream FEC coding)
pulling)
4K live LiveU 20Mbps 30fps 40Mbps, 10-2 (with 300–400Mbps, <50ms -95dBm 5dB
news which is 2× FEC coding) which is 15–
pushing the bit rate 10-5 (without 20× the bit rate
assurance) (burst increase)
above the base
station
CCTV Spring Cogent and 42Mbps 50fps 63Mbps, 10-2 (with SRT 700–800Mbps, <50ms -91dBm 8dB
Festival Sumavision which is assurance) which is 15–
Gala live 1.5× the bit 10-5 (without 20× the bit rate
stream rate assurance) (burst increase)
pushing above the base
station
Source: Huawei

Remarks:
• The preceding network coverage standards are based on 64T64R, 4:1 timeslot configuration, 100Mbps
bandwidth, and 1:1 5G/4G site deployment.

Currently, the 5G wireless network planning criteria are obtained based on the network with light load.
After the network data for network planning criteria is obtained based on the network with medium or heavy
load, this document will be further optimized.

Appendix
Further reading
Business Consulting Dept. (August 2018) 5G Service Requirement and Business Mode

User Experience Lab (eLab) (June 2019) Technical White Paper for 5G VR Service Experience Modeling

X Labs (June 2019) Huawei-CN-Cloud VR PI

Fixed Network Application Scenario Lab (iLab) (June 2018) Experience-driven Video Bearer Network White
Paper (2018)

X Labs (August 2019) 5G 4K Live Broadcast Service Analysis and Network Impact Report

© 2019 Informa PLC. All rights reserved. Informa Tech Ovum 23

OVUM CONSULTING
Ovum is a market-leading data, research, and consulting business focused on helping digital service providers,
technology companies, and enterprise decision-makers thrive in the connected digital economy.

Through our 150 analysts worldwide, we offer expert analysis and strategic insight across the IT, telecoms, and
media industries.

We create business advantage for our customers by providing actionable insight to support business planning,
product development, and go-to-market initiatives.

Our unique combination of authoritative data, market analysis, and vertical industry expertise is designed
to empower decision-making, helping our clients to profit from new technologies and capitalize on evolving
business models.

Ovum is part of Informa Tech, a B2B Information Services business serving the Technology, Media, and
Telecommunications sector. The Informa group is listed on the London Stock Exchange.

We hope that this analysis will help you make informed and imaginative business decisions. If you have
further requirements, Ovum's consulting team may be able to help your company identify future trends and
opportunities. Please contact us via

https://ovum.informa.com/contact/contact-us

consulting@ovum.com

https://www.linkedin.com/company/ovum/

www.twitter.com/Ovum