Mastering 5g network design, implementation, and operations

"We are living in an era where ultra-fast internet speed is not a want, but a necessity. As applications continue to evolve, they demand a reliable network with low latency and high speed. With the widespread commercial adoption of driverless cars, robotic factory floors, and AR/VR-based immersive sporting events, speed and reliability are becoming more crucial than ever before. Fortunately, the power of 5G technology enables all this and much more. This book helps you understand the fundamental building blocks that enable 5G technology. You’ll explore the unique aspects that make 5G capable of meeting high-quality demands, including technologies that back 5G, enhancements in the air interface, and packet core, which come together to create a network with unparalleled performance. As you advance, you’ll discover how to design and implement both 5G macro and private networks, while also learning about the various design and deployment options available and which option is best suited for specific use cases. After that, you’ll check out the operational and maintenance aspects of such networks and how 5G works together with fixed wireline and satellite technologies. By the end of this book, you’ll understand the theoretical and practical aspects of 5G, enabling you to use it as a handbook to establish a 5G network"

5G RAN deployment options

NR and NG-RAN features

The access network

The packet core network

The transport network Data center design 5G network concepts The AMF

Client/server – a REST principle applicable to 5G

Cacheable – a REST principle applicable to 5G

Stateless – a REST principle applicable to 5G NFs

A layered system – a REST principle applicable to 5G NFs

A uniform interface – a REST principle applicable to 5G NFs REST in 5G

The NR physical layer at a glance

Confinement of the spectrum with improved spectral efficiency

NR channel structure

The downlink reference and synchronization signals

Uplink reference and synchronization signals

NR cell definition

NR cell measurement signals

Waveform, numerology, and frame structure

5

5G Air Interface and Physical Layer Procedures – Part 2 Initial access and beam management

Initial access procedure

Beam-sweeping and initial access

Other remaining system information

Message 1, 2, 3, and 4 transmissions

CSI-RS and CSI reports

Channel coding schemes

Transport channel coding chains

Packet Core Procedures

PDU session establishment in 5G

PDU session establishment with 4G/N26/IWF 4G (E-UTRAN) initial attach

IMS 5G PDU session establishment

VoNR call flow

Wi-Fi attach

8

Voice over New Radio (VoNR) VoNR concepts and drivers

IMS network deployments growth

5G system service requirements

Ultra-high-definition audio

VoNR high-level architecture

5G NR access and core networks

VoNR call procedures

High-level call procedure activities

SIP signaling exchange

QoS flow establishment

QoS flow tear down

Interworking and voice call continuity Interworking with legacy voice networks

Packet switched handover from 5G to 4G

Single radio voice call continuity

5G network deployment architectures

NSA network architecture

SA network architecture

The different 5G deployment options defined by 3GPP

SA options

NSA options

A technical comparison of NSA versus SA

An economic comparison of NSA versus SA

4G to 5G migration strategies

The evolution of the EPC to 5GC

A comparison of the EPC and 5GC

Summary

5G Non-Standalone Networks

Introduction to NSA networks

Types of NSA networks

A deep dive into NSA networks

The architecture of the option 3 NSA network

5G NSA – option 3 call flows

Advantages and shortcomings

Summary

11

5G Standalone Networks

Benefits and use cases of a 5G SA network

5G SA network design fundamentals

Essential precursors

5G SA network components and end-to-end call flow Designing a macro network

Designing a non-public (private) network

5G network resilience and failure design

Network performance analysis

Summary

12

Managing O-RAN infrastructure

Using hybrid cloud

The emergence of the telecom cloud

Summary

13

5G Network Slicing

Understanding network slicing

Network slicing simplified

Network slicing challenges

Overcoming the network-slicing challenges Security considerations for network slicing Network slicing architecture and design 5G network slicing at work

Architecture and flow diagram

Use cases of network slicing

14

5G and Autonomous Vehicles

A background on CAVs and their concepts

Levels of automation

The 5G Automotive Association (5GAA)

The high-level architecture of AVs

Architecture for cooperative and intelligent transportation The role of cloud providers

Introduction to FMC in 5G

Session Initiation Protocol

Real-Time Transport Protocol

IP Multimedia Subsystem

Unlicensed Mobile Access

The Fixed-Mobile Convergence Alliance

The Broadband Forum

5G and Satellite Communications

Introduction to satellite communications

The Global Positioning System

Satellite communication in 5G

The role of regulatory bodies

Challenges to satellite communication

Why do we need satellite communication with 5G? The architecture of 5G and satellite communications

Use cases of 5G and satellite communication

Summary

17

Automation, Orchestration, and Testing

Automation infrastructure in 5G

Introduction to network orchestration

Auto scale-in and scale-out

Automatic failure recovery

Congestion control by configuration management

Implementation of network slicing

Integration planning in a multivendor environment

Monetization via AI and ML

Summary

Further reading

Appendix

Interview on 5G with Rob Tiffany, founder and CEO of Sustainable Logix

Interview on 5G with Sunil Dadlani, global CIO and CISO of Atlantic Health System

Interview on 5G with Aayush Bhatnagar, senior vice president of Reliance Jio Interview on 5G with ChatGPT

Index

Other Books You May Enjoy

Part 1:Introduction to 5G

This part of the book will provide an overview of 5G technologies You will develop a solidunderstanding of the fundamental building blocks that enable 5G technology The chapters inthis part will unravel the aspects that make 5G unique and capable of addressing high-qualitydemands in terms of bandwidth, latency, and quality of service, as well as the technologies thatback 5G, the enhancements to the air interface, and the packet core, which come together to light

up the network that is capable of this performance It will cover how to design and implement 5Gmacro and private networks and, at the same time, explain the various design anddeployment options available

This part of the book comprises the following chapters:

 Chapter 1 , Introduction to 5G

 Chapter 2 , End-to-End Architecture Components, Concepts, Security, and Transport

 Chapter 3 , Building Blocks – Cloud Native Infrastructure

 Chapter 4 , 5G Air Interface and Physical Layer Procedures – Part 1

 Chapter 5 , 5G Air Interface and Physical Layer Procedures – Part 2

 Chapter 6 , 5G Air Interface and Physical Layer Procedures – Part 3

 Chapter 7 , Packet Core Procedures

 Chapter 8 , Voice over New Radio (VoNR)

1

Introduction to 5G

5G is the fifth-generation technology standard for mobile cellular networks, which is the

successor to 4G networks This chapter introduces key aspects and methodologies of the 5G New Radio (NR), with a focus on the concepts and drivers It provides some basic understanding of the 5G NR and Next-Generation Radio Access Network (NG-RAN) and the end-to-end system

architecture at a high level Core network-related aspects will be evaluated in the upcomingchapters

Understanding the lessons of this chapter, mainly 5G concepts and drivers, and some keyfeatures of 5G NR is important to build the foundation for the upcoming chapters in the book

In this chapter, we will cover the following topics:

 5G concepts and drivers

 5G NR and NG-RAN

5G concepts and drivers

In this section, we will analyze key drivers for the need for 5G technology, key requirements,and the standardization of 5G.

Key drivers

Mobile technologies such as 3G, 4G, and 5G were initially governed by the International Mobile Telecommunications (IMT) requirements of the International Telecommunication Union – Radiocommunication (ITU-R) IMT-2000 was established by ITU-R with detailed

specifications for the first 3G deployments that took place around 2000 In early 2012,

ITU-R established the specifications of IMT Advanced for 4G wireless cellular technology Similarly,for the 5G technology, ITU-R defined IMT-2020

Figure 1.1 – ITU-R and the IMT technologies

IMT-2020 is the benchmarks and guidelines that the ITU-R has set down for what a 5G network

should be Today, organizations such as the 3rd Generation Partnership Project (3GPP)

are working toward fulfilling the requirements of IMT-2020 Within IMT-2020, there are three

use cases that are the main focus of 5G Those use cases include Enhanced Mobile Broadband (eMBB), Ultra-Reliable and Low Latency Communications (URLLC), and Massive Machine-Type Communications (mMTC) We will consider each of these in

turn

Enhanced mobile broadband

From 2G all the way through to 4G we have seen constant increases in the mobile broadbanddata rates that subscribers can expect to achieve 5G is no exception with its promise of eMBB

To be able to market 5G, some high data rates need to be provided to subscribers to show howcompetitive it is against 4G The headline data rates are roughly in the high hundreds of megabits

per second Certainly, 5G will deliver data rates that satisfy applications such as Augmented Reality (AR), ultra HD videos, or 3D applications.

Figure 1.2 – Evolution to ultra broadband

But certainly, with 5G, subscribers will typically experience data rates in the high hundreds ofmegabits per second

Ultra-reliable and low-latency communications

The second key use case is URLLC When we consider URLLC, we need to consider the factthat 5G will really be an enabler network So, we see a variety of different applications here that

might be able to use the 5G Core (5GC) Network Remote surgery, autonomous driving,

industrial control, and drone control are examples of applications that require low latencyand high reliability

Figure 1.3 – URLLC

URLLC has stringent requirements in terms of latency and reliability The latency for thenetwork is set at around 1 ms The network for certain applications needs to be super reliable aswell, with 99.999% (five 9s) reliability

Massive machine-type communications

The third key use case is mMTC and fundamentally, it is the cellular-Internet of Things (IoT).

Although we already had a cellular-IoT with earlier technologies, we see it again with 5G aswell There are numerous different IoT applications that can use the services of the 5Ginfrastructure The network must be super flexible and super adaptable from the 5G serviceproviders’ perspective The network needs to be able to provide exactly the correct requirementsfor the IoT applications that are using it Network function virtualization, network slicing, andedge computing came into prominence as the three key aspects of 5G These three aspects will

be examined later in the upcoming chapters

5G (IMT-2020) performance synopsis

The following table lists the enhancements of the minimum technical requirements of IMTAdvanced to IMT-2020:

Requirement Unit IMT Advanced IMT-2020

User-experienced data rate Mbits/s 10 100

Connection density devices/km2 105 106

Network energy efficiency bit/Joule 1x 100x

The following list expands on the preceding performance synopsis and key areas that serviceproviders today are moving toward:

 Peak data rate (Gbits/s): This is the peak throughput target that can be achieved by a

single user in the ideal radio conditions, and it is measured in Gigabits per second

 User-experienced data rate (Mbits/s): Shows the user-experienced throughput target,

which needs to be achieved by 95% of the users in dense urban areas This is the speedthe user will experience in the field

 Spectrum efficiency (bits/s/Hz): This is the number of bits per second per Hertz

achieved by 95% of users in the coverage area It indicates how efficiently thesubscribers can use the valuable radio spectrum

 Mobility (km/h): Shows how fast the subscribers can move while maintaining a specific

normalized traffic channel data rate

 Latency (ms): Represents the one-way delay between the time from when the source

sends an application packet to when the destination receives it

 Connection density (devices/km2): Shows how many devices can be supported per

kilometer squared This is something closely related to the cellular-IoT

 Network energy efficiency (bit/joule): Indicates how much energy is used in the

network to send a bit each time

 Area traffic capacity (Mbit/s/m2): How many megabits of information can be sent per

meter squared per second

5G standardization

Like many of the preceding technologies, 2G, 3G, and 4G, it is 3GPP that really defines thestandards 3GPP has defined the specifications for 5G, which are there to address the IMT-2020requirements Some of the techniques that were introduced in Release 14 were carried on toRelease 15 to be used as 5G techniques

5G was first standardized in Release 15 The first drop of Release 15 back in December 2017

provided a standard for service providers for Non-Standalone (NSA) operation within the

network However, Release 15 did not completely standardize every aspect of 5G Release 16and Release 17 includes further enhancements to 5G to provide full capability and address IMT-2020’s requirements

In terms of a timeline, back in 2017 and 2018, the earlier proprietary 5G systems started toappear; however, standardization was not complete at the time, so some NSA 5G networksstarted to emerge The period of 2019-2020 was really the time in which the first Phase-1deployments of standardized 5G based on 3GPP Release 15 commenced However, most of thenetworks that were deployed in 2019 centered around NSA operation, which is composed of 5G

RAN with Evolved Packet Core (EPC) Phase-1 deployments are only centered on eMBB

services It is Phase 2 where we see those additional two pillars of 5G, namely URLLC andmMTC

Phase-2 deployments are based on a combination of Release 15 and Release 16 features We seefull SA operations take place with the various features relating to URLLC and mMTC aswell as eMBB

In this section, we looked at the key drivers, performance synopsis, and standardization of 5G,which will help us understand the forces driving the technology we will be studying in this book.

We will now look at 5G NR and NG-RAN

5G NR and NG-RAN

Figure 1.4 shows the end-to-end architecture for the 5G system.

Figure 1.4 – 5G system high-level architecture

5G systems largely comprise the 5G NR, NG-RAN, and, finally, 5GC

NG-RAN architecture

Figure 1.5 shows key elements within the architecture of the NG-RAN The User

Equipment (UE) can be in the form of a mobile device, but it could also be in all manner of

different forms, as can be appreciated with the advent of IoT In order to provide RAN coverage,

there are gNBs, which stands for New Radio Node B A radio interface is needed to create

connectivity between the UE and the gNB That radio interface is called 5G Uu Notice also that

there is connectivity between the gNBs So, the Xn reference points allow these gNBs to

communicate with each other Finally, connectivity between gNBs and the core network is

needed, which is also achieved by the N reference points In the diagram, the N2 reference point

is providing the control plane flow, whereas the N3 reference point is used for the user plane,

which carries user traffic and user data

Figure 1.5 – NG-RAN architecture

The gNB is responsible for radio resource management and a big part of that is scheduling uplinkand downlink data onto the radio interface The gNB handles numerous different devices and itscell or even cells A single gNB may be in control of several different cells It is responsible forscheduling user data correctly onto the downlink or telling the UE when to transmit data on the

uplink The gNBs are also responsible for handovers The Xn reference point that sits between

the gNBs allows them to coordinate handovers between themselves Security is a key factor aswell, so the gNB will be involved in the security across the radio link as well as the UE Finally,the gNBs are responsible for dual connectivity, which will be examined later in this chapter

The UE is responsible for bidirectional data transfer via the 5G NR, ensuring a high quality ofservice So, the UE needs to be sure that the correct traffic is sent to the correct bearer By beingthe other end of the radio connection, the device is also responsible for security Finally, the UEneeds to support dual connectivity if it is being used in the network architecture

Tracking areas

Looking at the NG-RAN at a higher level, we can see that there are thousands of gNBs deployed

to provide that RAN coverage In any mobile communication system, the UE is not constantly

connected to the network The device will routinely be set to an IDLE state in order to preserve the battery life In the IDLE state, the UE will be conducting operations such as cell reselection.

But it will also periodically be listening to see whether it is being paged Any data coming intothe 5GC to be sent to the UE requires paging The network needs to know where that device is.However, the problem is that the cell reselection that takes place in an IDLE state is doneautonomously and the device does not keep the network updated as to which specific cell it’s in

If we need to page that UE, what we don’t want to have to do is page every single gNB in theNG-RAN So, consequently, the NG-RAN is broken down into tracking areas What’s crucial inthe system is that the UE might be autonomously making cell reselection when it is in

the IDLE mode The UE will keep the network updated as to which tracking area it is currently

in A tracking area is simply an administrative collection of gNBs and their associated radiocoverage, as depicted in the figure

The 5G Core Access and Mobility Management Function (AMF) in the 5GC keeps track of

the tracking area that the subscriber is currently in So, the subscriber’s UE will be required to

update the network any time it moves into a new tracking area The User Plane Function (UPF)

in the 5GC is responsible for the user plane data If user data comes into the UPF, it will informthe AMF and the AMF will page a specific tracking area instead of paging every gNB in thenetwork, since it already knows which tracking area the UE is in

Figure 1.6 – Tracking areas

The tracking areas effectively make paging much more efficient As soon as the UE connects tothe network, the AMF will know its cell ID But when it is IDLE, it just knows the location ofthe UE to the granularity of the tracking area Tracking area planning is all about making thepaging more efficient

5G RAN deployment options

Service providers will be in a transition phase as they move from 4G toward 5G networks Theycan’t just switch on the 5G network suddenly There are some strategy options for serviceproviders in terms of migration Now notice that in the diagram, there are 4G EPC and 5GCs

The first question is, will the service provider be deploying the 5GC or 5G Radio AccessNetwork first or will they both be deployed in parallel? Either way, there are variousconnectivity and deployment options available for the service providers to choose from It is notnecessarily the case of one or the other being used; it could be a mixture.

With the NSA approach, there is a 5G gNB, which supports dual connectivity to a 4G eNB, which stands for Evolved Node B So, the UE will be in communication with both RAN nodes

together It will have 4G radio connectivity to the eNB and 5G radio connectivity to the gNB.With the help of 5G radio connectivity, service providers can provide 5G services to theircustomers In this approach, notice that the control connectivity goes back into the 4G EPC sincethe eNB is the primary RAN device in this architecture So, this approach provides the benefits

of 5G gNB with 5G RAN coverage Alternatively, 5GC can also be utilized In this approach,

there is a gNB again and that gNB is in communication with a Next-Generation eNB (ng-eNB).

The main difference is that the ng-eNB connects to the 5GC But it’s a similar scenario wherebydual connectivity is used with the gNB as a secondary device and the ng-eNB as a primarydevice or master device and that control comes from the 5GC down to the ng-eNB Both theseoptions are NSA It depends on the service provider as to which approach they want to go for

The second option is SA, which is a pure 5G deployment The UE is using the gNB and thatgNB is connecting directly to the 5GC

Figure 1.7 – 5G Radio Access Network deployment options

Certainly, the NSA option is more straightforward to implement with a smaller investment.However, in the long term, network architectures and thus network deployments will be based onSA

NR and NG-RAN features

To meet the requirements of IMT-2020, such as coverage, capacity, and data rates, there aresome techniques and technologies that are employed Let’s look at these in detail

With this approach, a secondary RAN node works in parallel with the master RAN node toimprove the effective data rate that the UE can achieve To accomplish this, there must be acontrol and user plane connection between the secondary and master nodes Certainly, the datarate for the subscriber device can be significantly increased by taking this approach

The terminology used between the master RAN node and the secondary RAN node variesdepending on the network architecture and mobile technology In 5G, for instance, the masterand secondary RAN nodes could both be gNBs Alternatively, the master could be an ng-eNBand the secondary could be a gNB So, there are several different options available From theservice provider’s perspective, it is up to them how to deploy their dual connectivity solution

Figure 1.8 – Dual connectivity

Certainly, there are different approaches where a mixture of technologies is used between themaster and secondary RAN nodes, as can be found in 5G NSA deployments Fundamentally, byusing these two nodes together, better coverage and improved data rates can be provided for thesubscribers Note that the UE has also got to support dual connectivity This is particularlyimportant if, for example, the master RAN node is an eNB and the secondary RAN node is agNB, in which the device is then supporting 4G and 5G radios simultaneously.

Small cells

Small cells are nothing new; they are not a new technology Small cells have already been in usewith previous technologies for several years In this section, small cells will be examined in thecontext of dual connectivity in 5G

5G is set to really benefit from the deployment of small cells In the following scenario, smallcells are providing augmented indoor coverage where in-building penetration of the macro cell,especially in a high-frequency range, might be quite difficult Consequently, these small cells aredeployed within a building to improve indoor coverage and capacity in some cases However, for5G, indoor coverage is an important aspect of small cells and indeed outdoor small cells will

be routinely deployed

There are macro-level RAN nodes and small cells with the outdoor small cell deploymentapproach These macro-level RAN nodes act as the master RAN nodes, whereas the small cellsact as the secondary RAN node within the dual connectivity deployment

As the UE moves through the network, the blue line shows us the coverage that is experiencedfrom the macro-level RAN nodes The small cells provide data rate boosts to the UE as it movesthrough them Therefore, in a dense urban environment, these small cells, which may only have arange of hundreds of meters, can provide that augmented data rate boost to the network whileincreasing the overall capacity

Figure 1.9 – Macro coverage with small cells

These small cells can be set on the top of street poles, street furniture, lampposts, and so on.Small cells are a strong deployment option for service providers to really achieve the data ratesexpected for 5G.

Increased spectrum

In meeting the target of 100 megabits per second as a mean data rate, or potentially 20 Gbps as apeak data rate, it is essential for the service providers to have access to a more licensed spectrum.Consequently, the licensed spectrum bands that are being considered for 5G operation havebeen greatly increased

For 5G deployments, there are several bands in use They are below 1 GHz, 1-6 GHz, and 6-100GHz bands The below 1 GHz and 1-6 GHz bands are not new, and they are used by serviceproviders quite routinely 6-100 GHz is really the new band So, let’s explain why this newband is needed

Below 1 GHz is excellent for building penetration with wide area coverage The coverage ispotentially about tens of kilometers depending on the topography But the problem withoperating below 1 GHz is that there is not that much spectrum available for service providers,that is, there is limited spectrum availability So, what we need to do is start looking higher up inthe radio spectrum

1-6 GHz provides decent coverage, and there is also good spectrum availability 6-100 GHz islow range and only provides hundreds of meters of coverage, but there is greater amount ofspectrum available However, it is the key enabler for unlocking the stringent data raterequirements Service providers will be operating in much bigger bands higher up in

the licensed Radio Frequency (RF) spectrum, which is essential for providing those data rates.

As it has been with technologies that have come before and at the same time as 5G, ITU isresponsible for standardization and global harmonization of the RF spectrum At the WorldRadio Conference in November 2015, they already started to discuss and define some of theoperating bands for 5G, and at the World Radio Conference in October 2019, those bands wereset in place

Radio enhancements

5G NR means that the UE needs to be able to support 5G radio, and so too does the gNB Thereare many different tweaks that have been employed in NR protocols However, only two high-level aspects are going to be examined here

Fundamentally, one of the big changes is the employment of Cyclic Prefix Orthogonal Frequency Division Multiplexing (CP-OFDM) to provide greater deployment flexibility,

which adopts variable subcarrier spacing This helps service providers obtain much moreflexibility in terms of what kind of RAN coverage they want to provide Therefore, CP-OFDMallows them to address whether to deploy a small cell or a very large macro cell while allowing

them to be very flexible in terms of the frequency range that they can operate on Moreover,variable subcarrier spacing allows addressing specific latency requirements.

To address the requirements of the different IMT-2020 use cases, there will be a mixture ofdifferent cell types Some cells might require low latency and high frequency, whereas othersmight be long-range cells Therefore, to accommodate those different IMT-2020 requirements,such as latency and coverage, CP-OFDM has been introduced

Figure 1.10 – Radio enhancements

The other key area that has been introduced with 5G NR is the use of 256 Quadrature Amplitude Modulation (256QAM), which is already in use with LTE Advanced technology.

This is simply a radio modulation technique and effectively allows squeezing even more dataonto the radio carrier, hence increasing those data rates as appropriate However, to utilize256QAM the UE needs to be in a very good radio coverage environment

Beam forming

On the left-hand side of Figure 1.11, traditional antenna coverage is represented Traditionally,

the antennas might be sectored so they cover quite a large area Within that coverage area, theremight be fixed wireless access subscribers within the houses, mobile phones, or even fast-moving subscribers But the idea of traditional antennas is to cover a wide area

In 5G, Massive MIMO is set to be used A Massive MIMO antenna is a totally new

antenna design that comprises a huge number of radio frequency elements, in effectradio frequency antennas

One of the key advantages of Massive MIMO antennas is being able to form beams of radiofrequency energy This Massive MIMO array can provide much finer grain coverage If there is a

UE within that beam of coverage, the UE will get better general RAN coverage, which means thetechnologies such as 256QAM can be used to achieve the optimum data rate

Figure 1.11 – Beam forming with Massive MIMO

It is not only about creating that narrower beam of radio frequency energy; Massive MIMOantennas should be able to do that for numerous subscribers within the cell So, the beam willconstantly be flicking around to service different subscribers, which is called beam steering

Beam steering

Beam steering is the concept of beams following a UE in the network If the UE does not movetoo quickly, it can constantly provide feedback to the antenna so that the antenna can adjust thedirection of the beam that it is sending It is great for slow-moving UEs, while it is morechallenging for fast-moving UEs for the beam to keep track of them Basically, it depends on thespeed of the UE whether the beam steering is used or not

Figure 1.12 – Beam steering with Massive MIMO

Beam forming and beam steering are considered to be critical areas associated with 5G RANdeployments.

Cloud RAN

The final point related to 5G NR and NG-RAN techniques and technologies is the notion ofcloud RAN Essentially, cloud RAN sees the introduction of virtualization technologies to theradio access network With cloud RAN, each individual gNB has separate compute andprocessing resources The idea behind cloud RAN is to take that compute and storage capability

and move it to a Centralized Unit (CU) This CU is responsible for conducting the processing

and computing activities of all these gNBs So, the idea here is the compute capabilities arecentrally located in a data center and all that remains is down at the cell site are distributed units

The distributed units are the transmit and receive elements of the gNB This helps RANdeployment to be simplified and potentially cheaper The central controlling elementfundamentally is sending to the distributed units what they need to transmit The distributed unitsare simply responsible for transmitting, receiving, and exchanging that traffic with the CU

Figure 1.13 – Cloud RAN high-level architecture

The optical transmission links that connect the CU to the distributed unit in this approach arevery critical Fundamentally, these links must meet very low-latency requirements and they can

be over a kilometer long But in summary, here, the CU is simply using virtualizationtechnologies to control several distributed units

In this section, we analyzed the concepts of NG-RAN architecture, 5G RAN deployment options,and NR and NG-RAN features, which create the basis for the upcoming chapters, where we willexamine them in detail

Summary

5G standardization is being driven to meet the requirements of IMT-2020 We mentioned thethree pillars several times: eMBB, URLLC, and mMTC The theoretical maximum data rate for5G is 20 Gigabits per second, although the mean data rate for the subscriber is in the lowhundreds of megabits per second RAN latency requirements are around sub-1 ms.

3GPP standardization for 5G started with Release 15 and moved on to Release 16 In Release 17,

we also see enhancements taking place

Commercial deployments already started taking place in 2019 and onward into 2020 and 2021

We also analyzed the main components of the 5G system, including the 5G NR and NG-RAN Interms of the NG-RAN, there are two main elements: the user equipment and the gNB

We saw that tracking areas are designed to make paging more efficient by creating subgroups ofgNBs across the Radio Access Network

We discussed the different RAN deployment options available to service providers At a highlevel, they are called NSA and SA deployments

We also analyzed the NR and NG-RAN features, which gives us important knowledge forupcoming chapters First, we talked about dual connectivity It can be used to significantlyincrease a subscriber’s experienced data rate It is a key technology enabler for 5G Small cellsare closely related to dual connectivity We talked about how they can be deployed indoors oroutdoors With respect to licensed spectrum, we said it is essential to unlock an additionallicensed spectrum for the service provider so that they can operate the network at those high datarates

We also talked about the new air interface technologies We talked specifically about CP-OFDMand the use of 256QAM Beam forming antennas associated with the use of Massive MIMO will

be critical and with that, we also get beam steering Then, finally, we discussed Cloud RAN Wetalked about how it can provide efficiency and potential cost savings to the service provider ifthey choose to deploy it

In the next chapter, we will go over the end-to-end network architecture of 5G We will learnabout some concepts and the high-level components of access networks, packet core networks,and transport networks We will also understand how the quality of service is managed in 5Gnetworks

2

End-to-End Architecture Components, Concepts, Security, and Transport

In this chapter, we will learn about the various components that make up a 5G network We will

go into detail about the role of each of these components We will introduce some terms newlyintroduced in 5G and understand how they all fit in The topics discussed here will create a

foundation for understanding a 5G network and its various Network Functions (NFs) This

chapter will provide the knowledge base upon which we will build in upcoming chapters to beable to design and operate a commercial 5G network

We will go over the end-to-end architecture of 5G We will learn about what it takes to build a5G network from scratch We will learn about the various components that make up the network

We will also understand some concepts that are fundamental for understanding and building 5Gnetworks We will touch upon the transport and security aspects of 5G and also provide adecision chart that will help you with a jumpstart to build your own 5G network

In this chapter, we’ll work through the following topics:

A typical 5G network comprises three broad components that it uses to connect to the internet or

to applications Firstly, we have the radio component, also known as the access network Then,

we have the packet core network, which comprises various 5G NFs that we will discuss shortly.Thirdly, we see the transport network, which serves as the backbone of the 5G infrastructure andcarries the user and control traffic between the access network and the packet core network,

finally reaching the Application Functions (AFs) such as content servers or, simply put, the

internet

In Figure 2.1, we can see the 5G infrastructure components:

Figure 2.1 – 5G network overview

Let’s understand some of these components:

 ACCESS NETWORK: The access network comprises the radio components, basically the G-NodeB (gNB), also known as New Radio (NR) The access network is responsible for transmitting radio waves, which are received by the User Equipment (UE) This UE could include a mobile handset in a conventional use case, a

vehicle in V2X communications, or sensors and so on in advanced IoT use cases Theaccess network is hence the radio network and is restricted to communication over theair

 PACKET CORE NETWORK: The packet core network comprises control and User Plane Functions (UPFs) that are responsible for the management of the subscriber session, charging, data processing, the Quality of Service (QoS), and so on We will go

over each component shortly

 TRANSPORT NETWORK: The transport network is the backbone that connects the

access network with the packet core network and the internet Here, the word internet isloosely used – it can be any AF or content server, such as Netflix, a server with customcontent, or an IoT platform The most important facets of the transport network are that itshould be able to meet the latency and bandwidth requirements and also should be secureand not subject to attack by rogue elements

Now that we’ve understood the basic 5G network, let’s move on to looking at the high-levelcomponents

High-level components

Multiple components come together to bring up a 5G network In this section, we will learnabout the high-level components that are a must in creating and operating a 5G network 5Gprovides a lot of flexibility in the design of the network – we will also discuss the options as we

go over each high-level component

The access network

The access network comprises mobile base stations (gNB), which may be split up further

functionally and physically into the following components – a Centralized Unit (CU),

a Distributed Unit (DU), and a Radio Unit (RU) When there is a functional and physical split

of various components of a base station, it is called a distributed architecture, simply because asingle function of a gNB is now distributed among other smaller components

This split is very useful to support lower-latency use cases and helps serve mobile edgecomputing use cases Let us go over the components one by one:

 In the CU, the radio stack is decomposed to comprise Service Delivery Adaptation Protocol (SDAP) for NR, Packet Data Convergence Protocol (PDCP), which uses

packet processing components, including ciphering and header compression, and Radio Resource Control (RRC), which is responsible for controlling radio channels.

These components are not time-critical and hence they are good candidates for virtualization.The CU is generally deployed in remote near-edge data centers and is usually co-located withUPFs for low-latency applications The mid-haul transport network connects the CU to the DU

 The DU comprises the Radio Link Control (RLC), Medium Access Control (MAC), and Physical (PHY) layers, along with the brain of the radio base station, the scheduler.

The DU is a logical node hosting the RLC, MAC, and PHY layers of the gNB, and itsoperation is partly controlled by the CU The components in the DU are responsible forrate adaptation, channel coding, modulation, and scheduling radio resources among the

UEs in its coverage One DU supports one or multiple cells One cell is supported by one

DU The DU is connected to the CU by an F1 interface This is highly time-sensitive and

must operate within the guardrails of the latency requirements, which are in the order ofmilliseconds The DU can be located in near-edge data centers or can be collocated

with the Remote Units (RUs) for far-edge sites.

 The RU is the entity that is deployed at the cell sites, also known as Remote Radio Head (RRH) It connects to the DU through a CPRI or an eCPRI interface.

To summarize, the access network comprises the entities that control over-the-air communicationbetween the UE and the 5G network We will get into the details of this communication

in upcoming chapters

The packet core network

The packet core network in 5G comprises multiple independent NFs that have different roles andresponsibilities They are built on cloud-native development principles and operate following aproducer-consumer model Each of these NFs provides a set of services to different NFs andsimilarly, they consume services from other NFs Some of the most popular and essential NFs

are the Access Management Function (AMF), Session Management Function (SMF), UPF, Charging Function (CHF), Unified Data Manager (UDM), Policy Control Function (PCF), and so on These NFs provide mobility management, session management and

control, user plane and data processing, charging, subscription management, traffic policing, andQoS control and enforcement functions We will discuss each of them in the upcoming sections

The transport network

The transport network ties each of these network segments together and provides connectivitybetween them The transport network can be in various forms, starting from the front-haul –which is generally CPRI- or eCPRI-based and connects the RUs and the DU Next is the mid-haul, which connects the DU and the CU This is followed by the back-haul network, whichconnects the CU to the packet core network

It should be noted that it is of utmost importance that the transport network is fast: it should havelow latency and be reliable This ensures that it is failure-resistant and highly available It shouldhave the necessary bandwidth to support high data volumes, and it should also be secure and notprone to rogue attacks.

As you may have already guessed, the transport network will contain a web of networking gear,such as switches, routers, leased lines, microwaves, optical fibers, and so on Hence, based on therequirements, the desired design is selected The optical front-haul picks up gradually as itprovides reliable connectivity and increased bandwidth Microwaves can be used for relativelylong-distance connectivity Especially in rural and remote areas, solutions involve microwavelong-hauls Since they are cost-effective and easy to deploy, they are a contender as the preferredchoice over substitute optical fibers in some cases

In the preceding diagram, you may have noticed that a WAN is used to connect the back-haulnetwork This is because when a 5G solution is deployed in an entire country spanning thousands

of square miles, it is imperative for the service provider network to leverage optical fiber leasedlines to transport the user traffic from the operator’s 5G network to the internet or contentservers Hence, leased lines or a WAN are used to transport this data

Data center design

In the following diagram, we will go over a sample service provider 5G network template Here,

we can discuss the basic principles of network architecture and design, and we will go over theassumptions made and the options available that can be leveraged according to the use case athand:

Figure 2.2 – 5G data center design

To design a 5G network, it is of the utmost importance to first decide on what use cases need to

be supported The design of the network will change according to the latency, reliability, andbandwidth needed to support a particular use case In the example shown here, the use cases

supported are enhanced broadband, Ultra-Reliable Low Latency Communications (URLLC),

and the IoT, which means that this can be adapted to cover most applications

Let us focus on the main components of this design one by one – firstly, a far-edge site This isbasically the cell site in which the radio tower is erected As most of us will know, an entire

geographical area is broken down into cells in which radio base stations transmit The hexagonal

figures denote a combination of one or more cells, which can be connected to one or more RUs

Since this is close to the user accessing the network, it is known as far-edge.

It should be noted that for some deployment options, a far-edge site can host both the DU and the

RU In other cases, the DU can be located in a local data center close to a collection of sites toaggregate the data It is also possible to locate the DU, along with the CU, in a near-edge datacenter The near-edge data center can also host the UPF for delay-sensitive applications It is also

possible to collocate the CU, DU, and RU in a far-edge data center Hence, many differentlocation configurations are possible It is the operator’s choice to select the appropriatecombination The general idea is that closer proximity of these radio components to the UPF willreduce the latency and hence be more appropriate for delay-sensitive applications.

The CU is then connected to the mobile packet core by optical fiber The mobile packet core can

be centrally located It is also possible to regionalize the packet core in cases where thegeographical area of a country is large In this case, there can be multiple regions – for example,east, west, north, south, and central – that cater to the traffic and subscribers in proximity inorder to reduce latency Hence, the determining factor for the design choice is also thegeographical area of coverage Often, the connectivity of the CU to the packet core is via leasedlines, so it is important to ensure the lines are secure and use security protocols to avoidmalicious snooping and attacks that can cause denial-of-service

5G network concepts

Multiple NFs come together to make the 5G System (5GS) Each of these NFs has a designated

function and is connected to one or more other NFs to provide its services or consume servicesprovided by another NF

5G almost entirely ushers in the adaptation of virtualized infrastructure The NFs have been

designed to operate over Commercial Off-The-Shelf (COTS) hardware and are no longer

complicated or expensive These NFs are therefore virtualized They can be virtual based or use microservices-based or bare-metal-based software capable of running microservicesdirectly, bypassing the need for a VM The 3GPP standards on 5G have been written keeping thismicroservices-based virtualized infrastructure in mind Hence, 5G completes the cycle oftelecom virtualization that started with 4G and comes full circle with 5G

machine-3GPP has defined each of these connections between any two NFs, as shown by the reference

point architecture diagram in Figure 2.3 Each of the connections is named – this name

comprises the letter “N,” followed by a number to uniquely identify the connection betweenthe two NFs

We will go into the details of what role each of these NFs performs and how it interoperates withits neighboring NFs: