Spark New Zealand outlines 5G network intentions

Spark New Zealand Limited
ARBN 050 611 277 Spark City, 167 Victoria Street West, Private Bag 92028, Auckland, New Zealand


MARKET RELEASE
9 August 2018

Spark New Zealand outlines 5G network intentions

Spark New Zealand has published today a briefing paper that outlines how it is on track
to start providing 5G services to New Zealand consumers and businesses from 2020.
5G is the fifth generation of wireless communications technology, which is starting to be
rolled out around the world.
The briefing paper aims to inform investors of Spark’s 5G intentions, help customers and
stakeholders understand more about 5G, and address key considerations for
policymakers.
Managing Director Simon Moutter said Spark’s technical and network planning for 5G is
advancing after successfully conducting outdoor and indoor trials earlier this year.
“We are undertaking detailed planning to ‘map’ expected 5G cell site densities in New
Zealand and, as a result of this planning (and the learnings we have taken from our 5G
testing), we are forming a good understanding of how many new sites we will need for
5G, and where. We have already begun a build programme to increase the number of
cell sites in our existing mobile network – which will enable us to meet near-term
capacity demand as well as lay the groundwork for network densification required for
5G.”
Mr Moutter said 5G will enable Spark to provide additional capacity at a lower
incremental unit cost than under 4G and 4.5G: “This means that once 5G is available to
deploy, we will have a strong commercial incentive to rapidly build 5G network capability
as the primary means of keeping ahead of growing customer demand for more data at
faster speeds.”
As Spark’s network spend will be diverted from 4G capacity expansion into 5G as soon
as the necessary spectrum is available, Spark expects to fund 5G network development
(excluding spectrum and any material move towards widespread rollout of new cell sites
using high frequency mmWave band spectrum) within its existing capital expenditure
envelope of 11%-12% of revenues. By 2020, Spark expects its wireless-network specific
capex to be between 25%-35% of Spark’s overall capital expenditure envelope, up from
25% in the year ended 30 June 2017.

Spark New Zealand Limited
ARBN 050 611 277 Spark City, 167 Victoria Street West, Private Bag 92028, Auckland, New Zealand


In late 2018, Spark will launch a 5G Innovation Lab in Auckland’s Wynyard Quarter
Innovation Precinct that will allow partner companies to test and develop applications
over a pre-commercial 5G network.

Mr Moutter noted Spark is already making decisions that are contingent on securing
additional 5G spectrum and is having to make those decisions in the absence of any
clear government policy on when that spectrum will be available or in what bands.
He said the allocation processes for the two most likely spectrum bands – mid frequency
C-band and high frequency mmWave band - should be completed as soon as possible,
to ensure 5G services can be delivered in time for the 2020-21 America’s Cup in
Auckland as an international showcase opportunity.
In addition to these bands, low frequency spectrum (below 1000MHz) will be required to
deliver 5G services on a pervasive basis into rural areas (outside of small provincial
towns). The Government’s current work to define 600MHz spectrum as a band for
potential 5G use should continue at pace.
Mr Moutter said it was important for policymakers to recognise 5G is not a standalone
technology or solution. It will operate together with previous generations of wireless
technology and will be deployed as an overlay of existing network infrastructure.
Therefore, policy settings need to support network operators having control over the
evolution of their wireless networks.
The current competitive market model, in which multiple wireless network operators
compete against one another to grow their customer bases through product and service
innovation and pricing, represents a good blueprint for the way 5G can be rolled out in
New Zealand and would provide for more investment predictability and certainty over the
coming decade, Mr Moutter commented.
ENDS

For media queries, please contact:
Andrew Pirie
General Manager Corporate Relations
+64 (0) 27 555 0275

For investor relations queries, please contact:
Dean Werder
General Manager Finance and Business Performance
+64 (0) 27 259 7176

5G
The evolution
towards a
revolution
BRIEFING PAPER
Augus t 2018

Contents
PAGE 4 Executive summary
PAGE 6 Snapshot – customers, investors and policymakers
PAGE 8 5G – an overview
PAG E 11 What 5G will mean for New Zealanders
PAGE 15 Spark’s path towards 5G
PAGE 18 The Network
PAGE 22 Spectrum
PAGE 26 Network investment
PAGE 27 5G and competitive markets
PAGE 30 Beyond 5G
5G. The evolution towards a revolution.2

Introduction
Spark’s purpose is to help all of New Zealand win big in a digital world.
That means continuing our business
transformation to improve customer
experience, leading the introduction of
new technology in collaboration with our
business and industry partners, launching
world-leading new products and services,
and engaging with policymakers and
community stakeholders on how we can
ensure all New Zealanders benefit from and
thrive in a digital economy.
We’ve made big strides already. In recent
years, we’ve re-engineered our information
technology systems, launched new services
and partnerships, re-energised our culture,
and deepened our engagement with our
customers.
Spark’s ongoing success requires us to meet
the needs of our customers by adopting new
technologies that let them communicate
and do business - in New Zealand and on
the global stage. 5G - the fifth generation of
wireless technology - is a key element of that.
Just like previous generations, 5G will be a
step-change as businesses, entrepreneurs,
innovators and everyday users explore and
develop its potential. Its deployment will be
critical to our national infrastructure.
In this paper, we provide an update of 5G
developments in New Zealand and overseas,
we discuss what 5G will mean for our
customers and our investors, and we outline
Spark’s roadmap towards delivering 5G
services from 2020. We also highlight the key
considerations for policymakers to address if
New Zealand is to achieve rapid adoption of
5G that benefits consumers, businesses and
the broader economy.
5G takes us into the next technology wave.
It’s going to be an evolution that in time will
become a revolution.
5G. The evolution towards a revolution.3

Executive Summary
Spark is leading New Zealand into a new digital era that will transform the way we
communicate, the way we do business and ultimately the way we live.
The New Zealand Government has a goal to
support the development of the country’s fast
growing digital economy and enable New
Zealand to become a leading digital nation: “a
nation with a thriving digital sector, where our
businesses, people and government are all using
digital technology to drive innovation, improve
productivity, and enhance the quality of life for all
New Zealanders.”
1

Spark is committed to taking a leadership role
in helping New Zealand achieve this target. Our
purpose is to help all of New Zealand win big in
a digital world. With that purpose in mind, we
design and launch innovative new digital services
– such as our unbeatable unlimited mobile plan,
the 1 gigabyte per day free Wi-Fi we give to our
mobile customers, our recently-launched online
marketplace WeDo that connects consumers and
tradespeople in their local area, our Lightbox
TV service and our online streaming service for
Rugby World Cup 2019. Our purpose also drives
our planning for and investment in world-leading
digital technology and networks.
The next generation of wireless technology is
known as 5G and it’s now just around the corner.
We will build New Zealand’s best 5G network and
we’ve already started laying the groundwork for
that network so that, once 5G spectrum is made
available, we are in position to build our network
and launch 5G services in short order.
5G will start as an overlay of existing 4G and
4.5G networks.
Planning for our 5G network started some time
ago. In 2016, Spark was the first operator in New
Zealand to deploy 4.5G technology, as a pathway
towards 5G. During 2018, we have accelerated
that preparatory work and launched an extensive
cell site deployment programme that will increase
the density of our mobile network in preparation
for 5G. And in March, we conducted New
Zealand’s first ever live 5G tests in Wellington
and Auckland, achieving speeds of 9 gigabits per
second (Gbps) outdoors and 18 Gbps indoors –
hundreds of times faster than the typical speeds
experienced by most New Zealand wireless
device users today.
There is still much to be done to bring a
commercial 5G network to reality in New Zealand
– importantly, the Government is yet to allocate
5G spectrum. However, our 5G network planning
is well advanced and global device manufacturers
are gearing up to produce 5G-capable devices
f rom 2019.
The superior performance and cost-efficiencies
of 5G will encourage rapid network investment.
5G, like 4G and 3G before it, will deliver
customers higher speeds, more data, and better
performance, at lower incremental cost (per unit
of data) than previous generations of technology
could. And it will be critical in allowing Spark to
continue to keep up with our customers’ ever-
growing demand for wireless data. Data traffic
on Spark’s 4G network is almost doubling every
year and will start to outgrow our 4G network’s
sensible limits by around 2020-21. By that time,
it will make much more sense for us to invest our
capital in more efficient 5G wireless capacity than
in further expansions of 4G capacity.
5G technology creates the opportunity for a wide
variety of very different 5G services as and when
demand for them arrives.
5G will also open up a world of new possibilities
for New Zealanders that in time could transform
our lifestyles and our economy.
1
http://www.mbie.govt.nz/info-services/digital-economy
5G. The evolution towards a revolution.4

Rather than a single network with set
functionalities, 5G is better thought of as a
“network of networks”, with the potential for
multiple parallel networks known as “network
slices” - each designed to provide very different
performance characteristics - all sharing a
common physical network infrastructure.
At the same time as enabling revolutionary
technologies such as automated cars, virtual
reality services and advanced manufacturing that
rely on ultra-reliable real-time communication, 5G
will also allow smart cities to connect millions of
devices together to enable them to manage civic
infrastructure and utilities in a more efficient and
environmentally responsible way. Whereas today
the very different performance characteristics
of these use cases would require their own
standalone networks, in the future all of them
will be supported on a single shared 5G network
through the use of network slices.
We expect the 5G network (covering multiple
bands) will support multiple different services with
widely divergent connectivity requirements. This
will fundamentally change the economics of the
digital business models that many industries have
already identified as their future - but have yet
been unable to economically realise. We expect it
will allow them to push the bounds of those digital
business models even further.
Because of the extendable nature of 5G, we
will be able to deploy network slices delivering
particular performance characteristics or services
as and when commercial demand for those
services warrants. In each case, service-specific
equipment may be required to be deployed at
different parts of the network, and in some cases
network density - the number of cell sites required
to provide the required coverage - may need to
be increased, but these will simply be extensions
to the 5G network, able to be deployed as and
when there is sufficient demand for them.
While we will start by deploying a 5G network
that delivers superior performance and lower
incremental operating costs for our existing
mobile and wireless broadband services, that
same network will be readily extendable to
support the next generation of digital business
models and services as they arrive.
5G will initially deliver faster, better mobile and
wireless broadband services, but will be capable
of so much more.
Crucially, we do not expect the next generation
of digital business models and services will
be determined solely by telecommunications
network operators. Rather, they will be
conceptualised by the businesses and people
who live and work in the industries and
communities these new business models and
services will transform. The role of Spark will be
to collaborate with these businesses, ensuring
the right 5G capabilities (such as 5G network
slices and network densities) to support them are
built to demand. We want to put the challenge
out to all New Zealanders and to New Zealand
businesses to start that conceptualisation work
now, and to talk to us about it.
In the fourth quarter of 2018, we will launch
Spark’s 5G Innovation Lab in Auckland’s Wynyard
Quarter Innovation Precinct that will allow
companies and partners to test and develop
future 5G applications over a pre-commercial 5G
network. We’re actively looking for partners with
exciting 5G use-cases who can use the Innovation
Lab to refine and prove them.
5G. The evolution towards a revolution.5

SNAPSHOT
For Spark’s customers
• Spark is on track to start providing 5G services to consumers and businesses from
2020. Our technical and network planning is advancing; we have already conducted
successful trials and we will launch our 5G Innovation Lab in Auckland’s Wynyard
Quarter Innovation Precinct in late 2018 that will allow companies to test and develop
applications over a pre-commercial 5G network.
• With 5G, we will be able to offer customers much faster speeds and more data at
prices similar to what they pay today – as the technology allows us to deliver better
performance at a lower incremental cost per unit of data.
• Many of the services that 5G enables will involve massive growth in both the number
and type of connected wireless devices and sensors. In terms of consumer devices,
5G-capable wireless broadband modems are expected to be among the first available,
with production by the global manufacturers of 5G-capable smartphones increasing
rapidly from 2019 onwards.
• 5G services will initially be available mostly in built-up areas because the mid- and
high frequency spectrum bands that are likely to be first allocated by the Government
do not have the radio signal reach to make them feasible for use across large areas
of sparsely populated rural New Zealand (outside of small provincial towns). Once
low frequency spectrum bands are available for 5G, we will be looking to a range of
options, including network sharing, to address rapid network expansion as widely as
possible.
For Spark’s investors
• 5G will enable us to provide additional capacity at a lower incremental unit cost than
from continuing to expand 4G capacity.
• This means that once 5G is available, we will have a strong commercial incentive to
rapidly build 5G network capability as the primary means of keeping ahead of growing
customer demand for more data at faster speeds.
• We expect our 5G network development to be funded within Spark’s existing
capital expenditure envelope (excluding spectrum and any material move towards
widespread rollout of new cell sites using high frequency mmWave band spectrum), as
we divert spend from 4G capacity expansion into 5G as soon as we have the available
spectrum.
• Our investment profile will be moderate and within the current “normal” range for
capital expenditure of between 11%-12% of revenues.
• The current competitive market model, in which multiple wireless network operators
compete against one another to grow their customer bases through product and
service innovation and pricing, represents a good blueprint for the way 5G can be
rolled out in New Zealand. This model, if continued to be adopted by Government, will
provide for more investment predictability and certainty over the coming decade.
5G. The evolution towards a revolution.6

For policymakers
• With 5G, Spark will enable significant benefits for consumers and businesses, which in turn will deliver wider
benefits to New Zealand, enabling the country to maximise opportunities from global technology advances.
• A competitive market model involving existing wireless network operators and potentially new entrants is
the best way to ensure New Zealand consumers and businesses benefit from 5G.
• Future services will be increasingly complex and fast changing, so it is important to ensure both policy and
regulation allows for flexibility and encourages the rapid introduction and adoption of new services, for the
benefit of consumers.
• 5G services are not a standalone technology or solution. As with 4G/4.5G, it will operate together with and
will be deployed as an overlay of existing wireless network infrastructure. Policy settings need to support
network operators having control over the evolution of their wireless networks.
• Mid frequency spectrum allocated for 5G, in what is known as the C-band, needs to be available to
operators in sufficiently large blocks (of at least 80MHz, ideally 100MHz) to ensure they can build 5G
networks exceeding current 4G performance. Speeds and performance that materially exceed 4G will
require the use of much larger blocks (minimum of 400MHz, ideally two blocks totalling 800MHz) of high
frequency mmWave band spectrum.
• The C-band and mmWave band spectrum allocation processes should be completed as soon as
possible, to ensure 5G services can be delivered in time for the 2020-21 America’s Cup in Auckland as an
international showcase opportunity.
• Regardless of the operator, low frequency spectrum will be required to deliver 5G services on a pervasive
basis into rural areas (outside of small provincial towns). The current work to define 600MHz spectrum as a
band applicable to cellular networks should continue at pace.
• Innovative transport solutions will become increasingly important as there are more cell site locations
requiring high-speed connections to the network core. 5G requires new solutions for transport, using a mix
of microwave- and fibre-based technologies. Regulatory support may be required to ensure transport links
controlled by third parties are available on fair commercial terms to wireless network operators.
SNAPSHOT
5G. The evolution towards a revolution.7

5G – an overview
5G is the fifth generation of wireless communications technology and it’s starting to
be rolled out around the world. It’s the key to a worldwide wireless-centric technology
revolution that will develop over the coming decade.
5G will enable:
• Greater speed and capacity – the ability
to move more data, faster without a
corresponding increase in costs – as the
technology allows wireless network operators
to deliver better performance at a lower
incremental cost per unit of data.
oTypical speed improvements of up to 10
times faster than today’s experience, and
peak speed improvement from close to
1 Gbps possible in a 4.5G network to 10
Gbps.
o10 Gbps speeds are only possible using
high frequencies (i.e. mmWave band)
with large bandwidths of up to 800MHz
or more; when using mid frequencies (i.e.
C-band) peak speeds are typically 1-2
Gbps.
oBy comparison, the fastest speeds
available as of today for retail fixed
broadband services via the current
specifications of the ultrafast broadband
(UFB) fibre network is 1 Gbps.
• Lower latency – less delay/greater
responsiveness enabling real-time services to
be delivered.
oLatency improvements to 10 milliseconds
(ms) and potentially down to 1ms
from a typical 50ms today. This allows
extreme network responsiveness and
will eventually enable mass uptake of
augmented reality (AR) and virtual reality
(VR), as well as support mission-critical
applications for industry. Latency is the
delay that occurs when transmitting digital
data – anything less than 10ms is generally
undetectable to the human eye.
• The ability to connect many devices at once
– sensors and smart devices that comprise the
Internet of Things (IoT).
oEven today’s 4G networks are limited in
the number of devices they can connect to
simultaneously, because these networks
were designed to support population
densities. But we know that in the future
connected “things” – devices ranging from
fridges to streetlights to farm gates – will
far outnumber connected people, so 5G
technology has been designed to support
connected device densities of up to 1
million devices per square kilometre.
• Network slicing – tailoring the network for
specific uses.
oThis is, perhaps, the most transformative
feature of 5G technology: the ability to
tailor the network in accordance to the
performance requirements of a service
by virtualising functions and moving them
closer to the customer. The performance
requirements of, say, a connected
autonomous vehicle (ultra-reliable real-
time connectivity 24/7) are very different
to those of a smart parking sensor network
(low power, non-real time connectivity of
thousands of similar devices that will use
very small amounts of network capacity
infrequently and at random times). In the
early stages of 5G, network slicing will
be enabled by virtualising key network
elements, a journey Spark has already
started. But as it develops, the network
functions themselves will be virtualised
and located in the right place within the
network for the performance demands of
the customer’s service.
• Edge Computing – Taking more of the
network processing functions to the ‘edges’ of
the network.
oThis ensures network functions get the
bandwidth and low latency required
for key 5G services, by moving these
functions closer to the cell sites
supporting customer devices. But it
requires new network configurations
including more high-quality fibre
connections to these edge functions.
5G. The evolution towards a revolution.8

5G will take us from a world of connecting people to each other and to the internet, to a world of
connecting almost anything.
There are several technology realities that will influence the rollout of 5G. These include:
• As with any new generation of technology,
5G will initially be delivered as an overlay of
existing network infrastructure alongside
3G, 4G and 4.5G services – not as a distinct,
standalone network. This will enable 5G to be
deployed on a geographic basis as and where
traffic demand requires it, with customers
having consistent 4G coverage where 5G
coverage does not yet exist.
• The two spectrum bands that are likely to be
initially allocated for 5G are the mid frequency
C-band and the high frequency mmWave
band.
• Spark’s testing of C-band spectrum shows it
has similar signal propagation characteristics
(in terms of reach, penetration, etc) as the
1800MHz spectrum used for 4G. For this
reason, 5G services using C-band spectrum
initially won’t require a large number of new
standalone urban cell sites because much of
the cell site densification required is already
being undertaken (or will be in our plans) as
part of network capacity expansion over the
next few years.
• Although 5G services using mmWave band
spectrum (which enables extremely fast data
speeds over short distances) will require new
infrastructure in dense traffic hotspots, these
will be “micro-sites” not too much larger than
what are used today for commercial Wi-Fi
services. Most likely, these micro-sites will
be located in a hub and spoke configuration
around a macro-site that uses C-band
spectrum.
• C-band and mmWave band are not suitable
for pervasive coverage across rural New
Zealand (beyond small provincial towns),
because neither band has the radio signal
reach to cover large, sparsely populated
areas. Low-frequency spectrum bands (below
1000MHz) are needed to provide widespread
rural coverage – as occurs today with Spark’s
4G (700MHz band) and 3G (850MHz band)
services. For this reason, Spark has submitted
to the Government that 600MHz band
spectrum be made available for 5G (for more
details, see section on Spectrum).
Introduced to NZ in 1987.
Provided basic voice services using analogue transmission.
1
9
8
7
1
9
9
0
s
2
0
0
1
2
0
1
3
2
0
1
6
Made the switch to digital standards and introduced the
short-messaging service (texting). Spark used the CDMA
variant of this which was turned off in 2012.
Launched by Spark in 2001 (then upgraded in 2004 and 2009),
introduced data services.
Introduced by Spark in 2013, supported improved mobile broadband
and greater data speeds.
Launched by Spark in May 2016, extended the capabilities of 4G to
provide even more capacity and speed. Spark was one of the world’s
first wireless operators to launch 4.5G services, achieving peak data
speeds in excess of 1 Gbps.
Each generation of wireless technology has seen major advances over the previous iteration
4.5G
4G
3G
2G
1G
5G. The evolution towards a revolution.9

A growing number of markets around the world are pushing ahead with 5G
network deployment.
The USA, Britain and South Korea are
targeting commercial deployment of 5G
in 2019, while Canada, China, France,
Germany, Japan and Russia are aiming for
2020. In the USA, AT&T is expected to be
the first company with a mobile-oriented
5G network later in 2018, although it is not
clear what handsets will be able to use the
network immediately.
On average, in most countries, consumers
who are aware of 5G see themselves using
it within 30 months of its launch. Based
on consumer awareness and interest,
China and the United States might lead 5G
consumer adoption.
Surveys show overseas customer
expectations for 5G services are that most
will go mainstream within three to four
years of 5G launch. The leading initial
requirements from consumers for 5G are
better performance, coverage and pricing.
The global timeline
2012Initial research into 5G
2016Initial pre-standards testing
The first modem chip is released
20173GPP 5G Pre-Release standard for the 5G radio network is finalised
20183GPP Release 15 – 5G standards finalised
5G service demonstrations including at the PyeongChang Winter Olympics
Early networks are launched in Qatar and Finland. These are fixed wireless,
not mobile, networks using 5G over C-band
The global standard for the 5G core is finalised
Verizon, in the USA, has announced the launch of a 5G fixed wireless service
using mmWave in four cities late in the year
2019Qualcomm expected to build 5G Snap Dragon chip set for mmWave modem
5G-capable handsets are expected to be available
Some early mobile-oriented 5G networks expected to be launched:
AT&T, Verizon and T-Mobile in the USA all expect to have networks for 5G
smartphones available
World Radiocommunication Conference 2019 (WRC19) to standardise the
global use of mmWave frequencies
ITU to approve IMT-2020 standards (defining the requirements of 5G systems
as they relate to network operation, software virtualisation and fixed-mobile
convergence).
20205G networks expected to be launched in several more countries
20255G networks expected to be mainstream
5G. The evolution towards a revolution.10

What 5G will mean
for New Zealanders
5G will bring fibre-like speeds to mobile, allowing Spark to deliver wireless performance
that’s better, faster and with the capacity to meet projected demand.
It will allow millions of machines to communicate
with each other, and eventually it will allow complex
procedures to be performed remotely.
Applications that are already occurring in New
Zealand with 4G and 4.5G will be “supercharged”
by the superior performance, greater reliability
and lower incremental operating costs of 5G, as
they become commercially feasible and used more
extensively, and overcome the barriers of adoption
that currently exist due to network performance
constraints.
Features enabled by 5G that are expected to prove
attractive to consumers include the prospect of
“always on” connectivity (negating the need to
log on to Wi-Fi hotspots), faster downloading
of content, better-quality video calls and instant
access to cloud services and business productivity
tools wherever there is mobile coverage.
Key applications are expected to include:
Enhanced mobile broadband (eMBB): to meet
growing consumer demand for higher-definition
(e.g. 4K) video, information and social media
services, such as
• Wireless broadband that delivers fibre-like
broadband performance to homes and
business
• Mobile services that deliver unlimited data
plans and fibre-like speeds to mobile devices
• Richer media content, as 4K video becomes
commonplace (and the expectation of 8K being
around the corner).
Enhanced Machine Type Communications (eMTC):
to support connections and communication
between tens of millions of connected devices to
enable digital services that can help New Zealand
industries and homes become more efficient,
such as
• Smart city services - smart lighting and
smart energy, water and wastewater network
management services
• Smart home services - connected appliances,
smart home security services, and remote
control and operation services
• Real-time freight tracking and logistics
management services
• Stock, pasture, water, effluent and
environmental management services
• Wearables and tracking of children or elderly
• Health and safety monitoring and management
services
Ultra-reliable and low-latency communications
(URLLC): Supporting near-instantaneous
communications between connected devices
to support complex and integrated multi-user
networks and services, such as
• VR and AR industrial and entertainment services
• Remote operation of health, educational and
industrial equipment
• Autonomous vehicles and intelligent transport
systems, for example in high-risk zones like
airports or for urban transport routes
• Mission-critical applications, such as remote
surgery
5G. The evolution towards a revolution.11

One of the early applications of 5G for New
Zealanders is likely to be an expanded range of
wireless broadband products. Although wireless
broadband already provides a valuable alternative to
fixed broadband delivered via fibre cable or copper
wire, the limitations of 4G mean it is best suited for
homes and businesses with modest internet usage.
For instance, Spark wireless broadband is currently
offered with a maximum monthly data plan of 120 GB
for most customers, with 240 GB plans only offered
in selected areas where there is greater network
capacity. The improved capacity and performance
characteristics of 5G should encourage more
innovation of wireless broadband services – for
example, new pricing models (such as pay-as-you-
go) and new solutions for customers with special
use cases, such as boats on the move, or taking the
broadband modem to a holiday home.
But as we have seen with previous generations of
wireless technology, 5G will be just the starting
point of change for consumers.
Businesses, entrepreneurs and technology partners
will develop new applications that leverage the new
network capabilities and create services that are
compelling for users.
Although many applications for 5G are known today,
previous wireless technology step changes have
shown that other applications are likely to emerge
that we haven’t imagined yet. For instance, when
Spark launched its 4G network in 2013, services such
as wireless home broadband and unlimited mobile
plans were not on our immediate roadmap, but
both have since become significant new products
for our customers – driven by the adoption of media
entertainment services like Netflix, Lightbox and
Spotify and the mass uptake of smartphones.
A key learning from the 4G experience is there
needs to be sufficient flexibility and dynamism in the
New Zealand market structure to enable wireless
network operators to innovate and invest in new
5G-enabled services as opportunities emerge. Then,
as customers, industry and innovators learn more
about 5G and its possibilities, service innovation will
hit the fast track.
We can expect to see the remote monitoring and
adjusting of medical devices and implants, smart,
dynamic management of urban transport networks
that improve traffic flow and reduce the investment
in building more new roads, and a wide range of
VR or AR applications that will enhance the viewing
experience for major sports or entertainment events
such as a rugby test match or a music concert, or
for a virtual guided tour of a new building from the
construction site.
Longer-term services include smart-metering
and smart-grid applications for the energy sector,
immersive gaming, better fleet management, and
dynamic track-and-trace services for the logistics
VIGILAIR
VigilAir is the brainchild of ASG Technologies Ltd, a subsidiary of TPT Group Holdings (NZ) Limited, an
innovative NZ Company and a technology partner of Spark. The VigilAir SaaS product interfaces with customer
electronic security systems allowing drones to respond to security events – the security guard of the future,
providing a faster, safer and more cost-effective response. An important part of the system is the artificial
intelligence video analytics that can be ‘taught’ to identify objects, behaviours and events. Conducting
analytics from a moving airborne device is challenging, but can be enabled by a high-quality, high-definition
video stream that provides a higher level of detail to the analytics engine. The bandwidth and low-latency
of 5G will allow for the development of new VigilAir based capabilities, that will increase the opportunity to
deploy the system across industries and countries.
5G. The evolution towards a revolution.12

OHMIO
Ohmio is a NZ-based manufacturer
and developer of autonomous
vehicles. Ohmio has been
developing autonomous vehicles
since 2015, with the first four-seater
concept vehicles, the Ohmio HOP,
being launched in Christchurch
in September 2017. Ohmio is due
to release its first commercial
20-person vehicle, the Ohmio LIFT, in
2018. Ohmio’s autonomous vehicle
platform relies on multiple modalities
for sensing and communication – this
is so the limitations of one modality
of sensor or communication medium
can be mitigated through the
presence of other sensors. Ohmio’s
partnership with Spark allows Ohmio
to integrate 5G communication into
its vehicle and infrastructure with the
benefits of higher throughput. As the
first autonomous vehicle in the world
to operate using 5G, Ohmio intends
to stay ahead of competitors in a
highly competitive space.
industry. Technicians in one country will be able to
operate systems in another, offering global business
opportunities for New Zealand companies, for
example, dairy processing workers in New Zealand
operating factory equipment in another country.
New Zealand lags behind its peers in planning for
smart cities. Improving this will require integrated
planning by central and local governments as
well as the telecommunications industry. Some
potential services, such as driverless cars, will require
significant developments in devices (in this case cars),
related infrastructure (road system technology) and
legislation, as well as the upgrades to the wireless
networks.
5G’s ability to serve very different use-cases over
a common physical infrastructure is arguably its
defining characteristic.
Despite these services having widely divergent
network performance requirements, each can be
served over a common extendable 5G network.
In each case, service-specific equipment may be
required to be deployed at different parts of the
network, and in some cases network density - the
number of cell-sites required to provide the required
coverage - may need to be increased. But these
will simply be extensions to the network, able to be
deployed as and when there is sufficient demand.
Initially, the 5G network we deploy will be designed
to deliver improved mobile and wireless broadband
services to our customers – more mobile data
faster and at lower incremental operating cost
per gigabyte. But as our 5G network grows, and
as customers, application designers, businesses,
and sectors of our economy become aware of the
potential capabilities of 5G, we will extend that
network through the addition of network slices
designed to deliver more sophisticated performance
characteristics.
In this paper we set out some possible ideas for
what those characteristics might look like, and what
digital services and new digital business models
they might support. But we know that in practice,
it will be other participants in the ecosystem who
define the exact nature of many of the services to be
delivered over Spark’s 5G network. For instance, the
transport industry will design the new digital services
that dictate what a network slice for vehicles should
look like. And the same will be true of every other
sector of our economy that 5G has the potential to
transform. We are excited by the possibilities, and we
are designing our 5G network so it will be ready to
support those sectors, business models and digital
services, when they arrive.
5G will increasingly be about devices beyond the
familiar smartphones.
5G. The evolution towards a revolution.13

The extra speed and capacity of 5G will enhance
users’ experience with smartphones, but this will be
a “faster in more places” evolution rather than some
fundamental change. Many of the wider benefits
of 5G for customers will show up only when new
kinds of devices are enabled with 5G end-to-end
capabilities.
To a large extent, the mainstream consumer
take-up of 5G in New Zealand will be influenced
by consumers upgrading their wireless devices
(smartphones, etc) to the latest 5G-capable models.
For smartphones, we expect that devices will be
available with 5G C-band support from early 2019.
Whether these will be suitable for the New Zealand
market is still unclear because the devices need to
have a range of bands supported within them that
align with the range of bands in Spark’s network. For
example, the United States’ use of the 700MHz band
is quite different to New Zealand, so a US-market
handset may not work here. Typically, the second
iteration of any new device offers a wider range of
band support, so by 2020 there is a good prospect
that compatible handsets can be sourced for the NZ
market.
Although the device replacement cycle for most
customers is typically every three years or so, our
experience with the transition from 3G to 4G was
that a significant number of customers decided to
accelerate their device upgrades to take advantage
of the better performance and services available
via 4G. This meant that the New Zealand market
adoption of 4G occurred significantly faster and at a
larger scale than was initially anticipated by most in
the industry. We suspect a similar scenario may play
out with the transition to 5G.
Mobile broadband devices (including wireless
broadband, dongles, Wi-Fi hotspots) are expected
to be easier to obtain because they do not require
the same range of band support. These may start to
be available from late this year. Spark will be looking
to provide 5G modems as part of our continued
wireless broadband offering.
More advanced 5G-specific devices are likely to be
several years away from mass-market availability.
While prototypes exist already, there will need to
be service support from major global operators
for these solutions to have the volumes to become
generally affordable. Spark is following these
developments. Even before widespread availability
of these services, Spark is developing the back-end
capabilities (such as network slicing, virtualisation
etc.) that will be required to support them.
SENSORIUM
Immersive live event broadcasting via 5G will
be an early winner for VR. Building a long-tail
of quality content will also break down one
of the main barriers to mainstream consumer
adoption. From New Zealand, SENSORIUM
has been quietly creating and enhancing a
robust production and distribution platform
for live immersive experiences and comprises
their own VR cameras, VR editing system and
distribution application. SENSORIUM can
now produce live broadcast content, and
package and release delayed coverage of
any live sporting event.
Their ease of production has brought the
company into talks to secure the rights to
produce immersive content for Mixed Martial
Arts – the fastest growing sport in the world
with a first proof of concept being shot in
September 2018.
5G. The evolution towards a revolution.14

Spark’s path to 5G
Spark has been a leader in bringing the latest wireless technologies to New Zealand
since we launched the country’s first cellular mobile services in 1987.
After moving to 2G technology in the early 1990s,
we converted our network to all-3G in 2009 – and
followed that in 2013 with the introduction of 4G
services.
From mid-2016, we began enhancing our network
through the progressive rollout of 4.5G – involving
a combination of technologies that improve
wireless speeds and capacity over conventional
4G. This was our pathway towards 5G: the 4.5G
technology we deployed in our network enabled
us to learn how massive bandwidth needed to
be catered for not just within the Radio Access
Network (RAN), but at the aggregation and
Network Core levels, something we need to
understand for 5G.
Each change Spark has made on its path to
becoming New Zealand’s wireless technology
leader has pushed us to go further, but our
key strategy remains the same - to have the
industry’s leading wireless network that delivers
high-quality, reliable services to customers at the
lowest operating cost, per unit of data delivered.
Like previous generations of wireless technology,
5G will transform the economics for wireless
network operators, significantly reducing the
operating cost per gigabyte of data transmitted
– allowing Spark to deliver more data to more
customers, faster and more efficiently than ever
before.
In that sense, we see 5G as an evolution of our
existing 4G and 4.5G services, that will initially
be deployed as an overlay on the existing cell
site network and then extended as demand for
new services arrives. As that demand eventuates,
we’ll need expanded networks, denser networks
and decentralised core switching and control
functions – what’s known as edge computing - to
take full advantage of the speed, power, lower
latency and capacity 5G will bring.
The most capital-intensive component of this
expansion programme will be densifying our
network - increasing the number of cell sites to
build capacity and improve coverage. We are
undertaking detailed planning to “map” expected
5G cell site densities in New Zealand and as a
result of this planning (and the learnings we have
taken from our 5G testing), we are forming a good
understanding of how many new sites we will
need for 5G, and where.
We have already begun a build programme to
increase the number of cell sites – which will
enable us to meet near-term capacity demand
as well as lay the groundwork for network-
densification required for 5G. The locations and
design of these new sites have been chosen with
future 5G requirements in mind, and most will
cater to high traffic areas in major urban centres.
As these are mostly “infill” sites to strengthen
capacity between existing cell sites, most will be
While not all devices are fully 4.5G-compatible, many of the latest phones support features
that can take advantage of the technology and operate between three and five-times faster
than 4G. Now, 4.5G is available through 38 cell sites from Auckland to Invercargill, bringing
our customers faster speeds and giving the network more capacity – providing an early
glimpse of the possibilities ahead with 5G.
Most new cell towers will be smaller and less obtrusive
5G. The evolution towards a revolution.15

smaller and less obtrusive than conventional cell
towers – all are being constructed as extensions to
existing street lighting or utility poles.
In March 2018, we conducted New Zealand’s
first live 5G test site in Wellington, achieving
speeds outdoors of up
to 9 Gbps using high-
frequency mmWave
spectrum. An indoor
trial in Auckland the
following month,
also using mmWave,
achieved speeds of
18.23 Gbps. Such
speeds are hundreds
of times faster than
the typical speeds
experienced by most
New Zealand wireless
device users today,
which tend to range in
the low 10s of megabits
per second. The trials
have allowed us to test,
in a real-world environment, the speeds, coverage
and the parameters of the spectrum we’ll be using
for 5G.
Spark also conducted trials using mid frequency
C-band spectrum to compare the coverage to
what is currently achieved with the 1800MHz
band spectrum Spark has in use today. 5G will
deliver very similar coverage to 4G service in the
1800MHz band.
We’re also starting to actively engage with
customers and potential customers of future 5G
services.
In the fourth quarter of 2018, we will launch
Spark’s 5G Innovation Lab in Auckland’s Wynyard
Quarter Innovation Precinct that will allow
companies to test and develop future applications
over a pre-commercial 5G network.
We chose the Wynyard Quarter because the area
will be the event hub when Team New Zealand
defend the America’s Cup, yachting’s most
prestigious trophy, over the summer of 2020-
21. The event’s international profile and strong
technology focus makes it an ideal platform for
New Zealand to showcase exciting innovations
and new capabilities enabled by 5G.
Although our 5G lab’s pre-commercial network
is expected to use spectrum obtained from
the Government under a temporary licence,
from a technical viewpoint it could become the
starting point for commercial 5G services once
The Spark 5G Innovation Lab in Auckland’s Wynyard Quarter will
feature a showcase area to educate our customers on what 5G
can do for their businesses. It will have an area to display some
key-use cases for immersive media experience, Internet of Things,
and artificial intelligence.

Providing early access to a pre-commercial 5G network through
our global relationships with leading equipment vendors like
Huawei, Cisco, and Nokia will give our local partners a competitive
boost, fast-tracking these businesses’ 5G developments.

The purpose-built collaboration zone will help enable Spark to
collaborate with 5G Lab local partners in new applications such as
autonomous vehicles, drones and virtual reality experiences.
5G. The evolution towards a revolution.16

spectrum is available. For New Zealand to take
full advantage of the America’s Cup opportunity,
we would want to have a 5G network operational
from mid-2020. To meet this timeframe, we
therefore need to have certainty in terms of
spectrum allocation by late 2019.
Our customers’ response to 4G reinforces the
importance to Spark of leading the transition to
5G.
In the five years since 4G was launched, traffic
on our wireless network has grown 25 times -
it’s almost doubled every year - and well over
100,000 home broadband customers are using
a modem connected to the nearest 4G cell site
instead of a fixed-line connection. More than 95%
of handsets used on the Spark network are now
smartphones and 93% of our data traffic is now
delivered via 4G.
New Zealanders lives are increasingly mobile, and
the way New Zealanders engage with and use
technology is increasingly mobile. People want
more freedom and to stay seamlessly connected
wherever they are. We expect this emerging
preference for wireless services to intensify in the
coming years, and to reward service providers
who can deliver effortless customer experiences
by wireless.
We will use 5G to deliver services that reflect our
customers’ preference for wireless delivery and
to deliver the effortless service experiences that
customers value, such as:
• Traditional mobile services but faster and
increasingly moving to unlimited data.
• Wireless broadband services that are simple
to set-up and move but are also faster and
moving to unlimited data.
• Virtual reality experiences that transport our
customers into stadium or arenas anywhere
on the globe from wherever they are.
• Connectivity for more of our customers things
that is simple for customers to manage – from
wearables to home appliances to vehicles.
Our 4G experience also tells us that customers
will find new ways to take advantage of 5G
technology. When we launched 4G in 2013, we
could not have predicted the sharp growth in
video traffic on our wireless network. Similarly,
we do not yet know the full range of 5G use-cases
that might emerge. But already we are aware of
a number of potential 5G use-cases that will be
transformative in a number of sectors of the New
Zealand economy.
As these transformational digital business models
and digital services eventuate, we know there
will be incremental value in delivering the 5G
connectivity that supports them, and further
value again in providing the data analytics and
management services that will underpin them,
because they will allow sectors of our economy
to realise significant cost efficiencies that those
sectors will value.
Our ambition is to have New Zealand’s best and
most extendable 5G mobile network in place
so when commercial demand for those digital
business models emerges, we are ready and
primed to extend that network in whatever ways
are necessary to enable them.
1,200
1,000
800
600
400
200
0
A
v
e
r
ag
e m
o
nt
h
l
y d
at
a per

c
o
n
nec
ti
o
n (
M
B
)
2009/102010/112011/122012/132013/142014/152015/162016/17
Mobile data consumption continues very strong growth
Source: 2017 Annual Telecommunications Monitoring Report, NZ Commerce Commission
5G. The evolution towards a revolution.17

The Network
Spark’s 5G services will be delivered over a network comprising cell sites connected
by fibre-optic transport or wireless transport links, an upgraded network core, edge
computing and a radio access network (RAN).
The evolution to a 5G network will be built in what
is termed a ‘non-standalone’ implementation. This
means 4G will continue to provide base coverage
to our customers, particularly at the connectivity
layer. Advanced services will be enabled with the
overlay of a 5G network. The overbuild will involve
moving to a distributed network where more
network functions are aggregated at points closer
to the cell sites being used by customers, with
high-quality transport connections back to the
main Spark core.
We expect the initial 5G network will focus on
using C-band spectrum – not only because this is
the first spectrum band likely to be allocated by
the Government, but because this mid-frequency
spectrum gives the best balance between
capacity and coverage. Spark’s testing of C-band
spectrum shows it has similar signal propagation
characteristics (in terms of reach, penetration,
etc) as the 1800MHz band spectrum used for
4G. For this reason, 5G services using C-band
spectrum initially won’t require a large number
of new standalone urban cell sites because much
of the cell site densification required is already
being undertaken (or will be in our plans) as part
of network capacity expansion over the next few
years.
In contrast, high frequency mmWave spectrum
enables extremely fast data speeds and high
capacity over short distances, which makes it
well suited for dense traffic hotspots. Within
these hotspots, we will need to build a significant
number of new cell sites – however, these will
be “micro-sites” not too much larger than what
are used today for commercial Wi-Fi services.
These microsites can be fixed to “street furniture”
Mobile edge and aggregationFixed edge and aggregation
Wireless
Backhaul
Fibre Backhaul
Fibre BB & Business Services
IOTVR/AR – connected carsWireless broadband
Wireless
Backhaul
Spark network
Spark core
5G4G
Virtual
Network slice
‘X’ haul transport supplied by:
• Chorus
• Other LFCs
• 3rd parties
• Spark
• Microwave
5G. The evolution towards a revolution.18

ComponentDescription5G Future
DeviceThe final step in providing the
service to the user. Most commonly a
handset.
Proliferation of device types including data
devices, various sensors, as well as macro-
devices (fridges, drones, cars) that have
wireless functionality embedded
SpectrumThe radio link from the device to the
cell site
New spectrum bands will provide
materially higher speeds and network
capacity
Radio Access
Network
(RAN)
Cell sites and the radio access and
associated processing components
Cell sites become progressively smaller
and more numerous, requiring changes
in deployment strategies and in the
architecture of the network
Transpor t
Links
In a 5G network, the transport links
between the RAN back into the core
of the network are called fronthaul,
midhaul and backhaul
Fibre will remain the primary means of
transport between cell sites, but low
latency demands mean transport to the
cell edge becomes more time-critical and
bandwidth much higher. Smaller sites
will require cheaper solutions and most
likely get aggregated at a distributed
aggregation hub. Wireless options will
become more common to feed back to
hub sites.
Network
Core
The central processing in the network
where services are defined and
users are authorised. There are
two broad functions: the ‘control
plane’ that provides functions
such as authentication and service
authorisation; and the ‘user plane’
that routes customer data traffic
and provides real-time charging for
services
With 5G, the two network core functions
need not be co-located as they are today.
The network will support Software Defined
Networking and Network Function
Virtualisation and allow Spark to support
multiple service categories within the
network core.
Distributed
Edge
The point where the cell sites
consolidate and connect back to the
network over transport solutions to
carry data.
5G will operate in a hub-and-spoke
environment. Mini data centres will house
virtualised network functions and act as
a consolidation point for several sites to
communication back to the core network.
Fibre will be key for linking these back to
the main Spark network.
5G. The evolution towards a revolution.19

such as lamp posts and power poles, and other
locations such as the sides of buildings, making
them much cheaper to build than a conventional
cell tower. Most likely, these micro-sites will be
located in a hub and spoke configuration around
a macro-site that uses C-band spectrum.
As our 5G network expands to encompass these
micro sites that will be required to provide dense
mmWave spectrum coverage, this will require
a change in fibre transport services and pricing
structures. Similarly, the cost and complexity
of installing these micro sites will need to be
orders of magnitude lower than that faced when
building cell sites today. As an industry we will
need to work with central and local government
to facilitate low-cost deployment of these lower-
impact, smaller-coverage sites ahead of demand
for them.
The core of the wireless network represents the
centralised processing that provides functions
such as authorisation, measurement of customer
data use and routing. The core will evolve towards
new technologies and some functions will be
“virtualised” so they can be moved closer to the
network edge. This will be crucial for low latency
services such as AR or VR. These functions will
increasingly occur closer to customers on generic
hardware - what is known as virtualisation. The
physical location of such processing will also
move outwards to more locations and closer to
the end users (edge computing).
The flexibility of 5G has been realised through the
ability to “virtualise” functions within the network
that enable the provision of tailored services for
specific customer needs or use cases. With 5G we
move away from the “one-size fits all” paradigm
of every previous generation, to one whereby
the network exhibits multiple personalities to
the services it enables. Different metrics (such
as bandwidth, latency and reliability) can be
presented to different services.
The 5G network core provides two broad
functions: a “control plane” function, which
oversees the request to connect by a device
and a “user plane” function, which allows the
information exchange between devices. These
could take place in different parts of the network.
For example, with time-critical services like virtual
reality or traffic control information, the user plane
can be located close to the device location while
the control plane can be in a data centre deep
within the network.
5G transport is different from 4G in that there
are three components – fronthaul, midhaul and
backhaul. These will use a mixture of radio and
2
For more information on the Rural Connectivity Group, refer to http://www.thercg.co.nz/
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
% o
f d
at
a t
r
a
f
fic o
n 4
G
S
e
p
t

2
0
1
3

—
No
v 2
0
1
3

—
J
a
n

2
0
1
4

—
Ma
r 2
0
1
4

—
Ma
y 2
0
1
4

—
Ju
l 2
0
1
4

—
S
e
p

2
0
1
4

—
No
v 2
0
1
4

—
J
a
n

2
0
1
5

—
Ma
r 2
0
1
5

—
Ma
y 2
0
1
5

—
Ju
l 2
0
1
5

—
S
e
p

2
0
1
5

—
No
v 2
0
1
5

—
J
a
n

2
0
1
6

—
Ma
r 2
0
1
6

—
Ma
y 2
0
1
6

—
Ju
l 2
0
1
6

—
S
e
p

2
0
1
6

—
No
v 2
0
1
6

—
J
a
n

2
0
1
7

—
Ma
r 2
0
1
7

—
Ma
y 2
0
1
7

—
Ju
l 2
0
1
7

—
S
e
p

2
0
1
7

—
No
v 2
0
1
7

—
J
a
n

2
0
1
8

—
Ma
r 2
0
1
8

—
Ma
y 2
0
1
8

—
Ju
l 2
0
1
8

—
3G
4G
% of data traffic on 4G vs 3G
4G take-up by Spark customers
Source: Spark NZ data
5G. The evolution towards a revolution.20

fibre technologies and, for the latter, Spark will use
a mix of its own fibre and third-party fibre. As sites
become smaller and more numerous over time, the
per-site transport costs become more significant.
Smaller sites will need much lower transport costs
either by way of unbundled access to existing
shared-fibre networks or by materially lower pricing
of uncontended multi-gigabit transport services.
Network sharing will be driven by the need to
meet customer demand as efficiently as possible.
For the most part, especially in cities and towns,
we envisage each wireless network operator will
build and operate its own network infrastructure as
occurs today with 3G and 4G services. Competition
between wireless networks has delivered
immeasurable benefits to New Zealand in the last
30 years and competition between 5G networks will
continue this trend.
In contrast to the wireless market, fixed line
telecommunications in New Zealand is dominated
by monopoly fibre and copper access providers,
with retailers offering their customers resold
connectivity to those networks. As a retailer of
both fixed and wireless telecommunications
services, Spark has experience operating within
both of these market structures. Network
competition between competing wireless
networks has permitted mobile retailers to offer
truly differentiated services and thus dynamic
competition and efficiencies to customers. That
encourages retailers such as Spark to invest in new
technologies and services, providing true choice to
consumers.
However, Spark also recognises that in certain
circumstances infrastructure sharing between
mobile network operators can deliver better
outcomes for our customers. Examples are in
highly-concentrated urban areas with high demand
where it is difficult to obtain the approvals needed
to add new cell sites or difficult to serve rural areas
where coverage by multiple mobile networks is
uneconomic.
Spark, Vodafone and 2degrees already share
wireless networks under the Rural Connectivity
Group (RCG)
2
to provide wireless coverage in
remote areas, and we already share cell towers and
some other network elements, as well as spectrum
and radio equipment in suburban and urban
areas where this is required. As 5G networks are
deployed we will continue to look for opportunities
to share infrastructure investment with other
wireless network operators where it makes
commercial sense to do so and where it can extend
5G coverage across New Zealand.
Sharing it is likely to make the most sense in
remote areas, where even cell sites using the
lowest-available frequencies don’t cover enough
customers. In these areas, an RCG-style shared
network build is likely. Infrastructure sharing
might also involve co-location of multiple wireless
network operators’ equipment on the same cell
site. This is like phase one of the Rural Broadband
Initiative (RBI), where Spark co-located on many
of the RBI sites built by Vodafone under its RBI
contract with the Government.
5G. The evolution towards a revolution.21

Spectrum
Spark is already making decisions that are contingent on securing additional 5G
spectrum and we are having to make those decisions in the absence of any clear policy
on when that spectrum will be available or in what bands.
Globally, C-band and mmWave are the most
referenced bands for initial 5G deployment. There
is enough certainty around them to have a good
level of confidence that they will be key parts of
the New Zealand 5G ecosystem in the near term.
C-band
C-band spectrum exists in frequencies between
approximately 3300MHz (3.3GHz) and 5000MHz.
In New Zealand, the Government is considering
allocating C-band spectrum from about 3400MHz
up to about 3800MHz. C-band is an extension
of the spectrum used by advanced 4G networks
today. The principal benefit it offers to wireless
operators and our customers is larger blocks of
spectrum than we have been able to use for 4G
and preceding technologies. A feature of each
new generation of wireless network technology
has been the ability to use larger blocks of
spectrum; in simplified terms, the more spectrum
a network can use the faster the data speeds on
that network will be. This means larger blocks of
spectrum than wireless network operators have
purchased in the past for 4G and 3G networks
will be needed to take full advantage of 5G
technologies.
Potential 5G bands identified by government
* The Government proposes that C-band (3400–3800 MHz) is top priority for 5G implementation, with mm-wave band (24–29 GHz) a high priority
(Ministry of Business, Innovation and Employment discussion document ‘Preparing for 5G in New Zealand’, March 2018).
Spark’s existing spectrum bands for 3G/4G/4.5G
Reach from cellsite
Speed & capacity
Slower
Shorter
Higher
600 MHz700 MHz850 MHz2100 MHz1800 MHz2300 MHz2600 MHz3400–3800
MHz (C-band)*
24–29 GHz
(mm band)*
5G. The evolution towards a revolution.22

There are currently two, and in the future there will be three, key
spectrum bands used on cellular networks.

Low frequency spectrum, below 1000MHz, is used to transmit
cellular signals over longer distances – 20 kilometres is often quite
achievable over flat terrain. This spectrum is used to provide wide-
area coverage and to provide in-fill between sites using higher
frequencies. Spark has rights to the 700MHz and 850MHz bands
already used for 3G and 4G services and we anticipate the need for
spectrum within, say, the 600MHz band to serve rural New Zealand.

Mid frequency spectrum, from 1000MHz (1GHz) to around
6000MHz, is used for the key capacity bands for today’s cellular
networks. As there is more spectrum available than in the low
frequency bands, more cellular traffic, in aggregate, can be carried.
This spectrum does not perform as well as low frequency bands in
rural areas, although in many cases a small provincial town might
have mid frequency spectrum on a cell site to carry the in-town
traffic – freeing up capacity on the low frequency spectrum to serve
the surrounding area. Spark has existing rights to 1800, 2100, 2300
and 2600MHz bands. In the New Zealand context, 5G C-band is
defined between about 3400MHz and 3800MHz.

High frequency spectrum, in the mmWave band above about
24GHz, is currently used for point-to-point wireless links and
connections to satellites. It has been identified as spectrum for the
expansion of 5G because it offers significantly greater capacity
than any of the bands currently in use in cellular networks. Its
disadvantage is that its reach (propagation) is materially lower than
existing bands, so deployment just on existing sites would result
in limited coverage. Therefore, it is expected that 5G networks will
need to increase their numbers of cell sites over time to provide
consistent mmWave coverage – but they will be much smaller, and
cheaper to build, than traditional cell sites.
Low
Mid
High
5G. The evolution towards a revolution.23

To capitalise properly on this potential, minimum
C-band spectrum blocks of 80MHz (and ideally
100MHz) should be available per operator
so initial 5G services can deliver data speeds
better than what 4G can currently achieve using
2300MHz band spectrum.
mmWave
The other key band for initial 5G use is referred to
as “mmWave”. In New Zealand, the Government is
considering allocating spectrum in the 24.25GHz
to 28.35GHz range initially, with some higher-
frequency mmWave bands likely to be made
available later. This spectrum band is much higher
frequency than any currently used in New Zealand
for cellular wireless networks and its physical
characteristics are quite different. The coverage
footprint per site is materially smaller, and the
ability to penetrate into buildings, other than
through windows, is quite limited. Conversely,
the high frequency means even larger spectrum
blocks are available - promising extremely high
data speeds when users can receive coverage.
To optimise these data speeds, spectrum blocks
in the mmWave band should be of at least
400MHz, with the opportunity for an operator
to acquire two blocks totalling 800MHz. Subject
to spectrum availability, Spark intends to start
installing smaller, localised cell sites suitable for
mmWave soon after C-band macro cell sites have
been deployed.
Spark believes the Government’s policy priority should be to make C-band and mmWave
spectrum available for purchase and use as quickly as possible, with block sizes of at least
80MHz and 400MHz respectively.
5G. The evolution towards a revolution.24

Environmental concerns
One of the concerns relating to cellular technology is the impact of electromagnetic fields (EMF)
generated by cell towers and related equipment.
In New Zealand these are governed by the New Zealand Standard NZS2772 recommended by the
Ministry of Health (MoH) and mandated under the Resource Management Act. The New Zealand
standard sets maximum public exposure levels that are more than 50-times lower than the recognised
threshold for established effects.
Spark designs all its mobile sites to comply with this standard and employs independent monitoring to
ensure ongoing compliance. Monitoring to date shows average EMF exposures are, in practice, a very
small fraction of the levels set in NZS2772.
It is expected that New Zealand and international agencies will continue to monitor the effects of 5G
deployments as networks are rolled out and to explicitly consider whether any change to the standard is
required. Spark’s deployment of 5G will follow the standards set by MoH and international bodies.
We see commercial use cases for each of these
bands in the near term and it’s imperative
New Zealand keeps pace with international 5G
deployments. Key policy decisions on this need to
be made in 2018 and spectrum auctions need to
occur during 2019.
Low frequency spectrum
Looking further ahead, the first low frequency
spectrum we expect to see released for 5G use
in New Zealand is in the 600MHz band, some of
which is currently used for broadcasting services.
Other spectrum bands at frequencies below
600MHz might also be considered in future for 5G,
as they would have similar signal reach that would
make them suitable for rural cell site deployment.
Although good service can be provided to
customers sufficiently close to a C-band or
mmWave spectrum cell site (including those in
small provincial towns), in most rural areas there
are often many customers too far from any site for
these bands to be effective. Spark is responding to
this in two ways:
• We have advocated for the release for
cellular use in New Zealand of 600MHz band
spectrum. This will have even better reach than
the present preferred 700MHz rural spectrum
band. T-Mobile in the USA is expected to offer
services using the 600MHz band from 2019.
• We have built a LoRa
4
network to provide
deep, wide-area coverage for sensors and
other IoT devices with applications in the rural
sector. This ensures we can efficiently deliver
certain IoT applications to the rural sector in
advance of a 5G network rollout.
Spark is already making decisions that depend
on securing additional 5G spectrum and we are
encouraging the Government to make clear policy
decisions on what spectrum will be made available,
and when, for 5G services in New Zealand.
To support the build of new infrastructure and equipment, changes may be required to
planning practices. We support the Ministry for the Environment initiative to develop
a utilities planning standard under the Resource Management Act
3
, setting out the
policies and rules for network utilities. These will potentially replace separate district plan
requirements.
3
The Planning Standard for Infrastructure and Roading

4
LoRa is a Low Power, Wide Area wireless network. It is not a 5G network, but can be used to supplement 5G network coverage for IoT devices.
5G. The evolution towards a revolution.25

Network Investment
User demand is growing rapidly on our 4G networks and we are continually adding
capacity - 4G spectrum carriers, sites and sectors - but this will take us only so far.
With our customers’ ever-growing thirst for data
services resulting in wireless network demand
roughly doubling each year, 5G will enable us to
provide additional capacity at a lower incremental
unit cost than continuing to expand 4G capacity
would.
This means that once 5G is available, we will have
a strong commercial incentive to rapidly build
5G network capability as the primary means of
keeping ahead of growing customer demand.
Based on current network traffic growth rates,
we anticipate that by 2020-21, 5G will be our
preferred solution for capacity augmentation – let
alone for the anticipated demand growth from
new applications made possible by 5G.
We expect our 5G network development to be
funded within Spark’s existing capital expenditure
envelope (excluding spectrum and any material
outlay required for a more widespread move to
mmWave), as we divert spend from 4G capacity
expansion into 5G as soon as we have the
necessary spectrum.
As Spark responds to demand we will be
investing just ahead of it. Cost efficiency that
will deliver ever-greater output with the same
investment inputs is the primary driver of early
5G deployment. By 2020, we expect our wireless-
network specific capex to be between 25%-35%
of Spark’s overall capex envelope. This implies
intended annual wireless network investment of
approximately $100m to $140m, compared with
an average of just over $100m for the past five
years. This excludes spectrum purchases and any
material move towards widespread rollout of new
cell sites using mmWave band spectrum. During
this period, we expect our total capex (excluding
spectrum) will remain in line with our desired
range of 11%-12% of revenues.
Revenues from wireless connectivity are expected
to increase only moderately over the first few
years of 5G deployment, in line with recent trends.
From about 2023 onwards, there is potential for
further upside as customer demand grows for
new capabilities, rather than just more capacity
and speed.
5G is expected to be deployed initially as an
evolution of existing 4G/4.5G networks relying
on the existing network to ensure comprehensive
coverage. Some services, such as voice and
emergency calling, rely on the existence of a 4G
network.
As further demand growth intensifies and new
use cases for 5G emerge, particularly those based
on URLLC and eMTC, Spark will continue to invest
in additional standalone capability and further
densify the access network to add capacity.
5G. The evolution towards a revolution.26

5G and
competitive markets
Wireless competition in New Zealand is vigorous and has served the country well.
Not only is the price of service packages
low but coverage is expensive to provide, so
the outcomes in this country are particularly
impressive.
This comes down to competition, which has
featured a constant trend of infrastructure
upgrades and innovations as the three wireless
network operators have vied for market position.
The result has seen significant benefits to
consumers and the same competitive approach
will best deliver efficient 5G infrastructure and
future technology upgrades for New Zealanders.
In addition to delivering great outcomes for
consumers, a competitive market structure in
New Zealand is the most cost-effective solution
for taxpayers – because the private owners of
the wireless network operators assume the vast
majority of investment required to build new
networks and expand capacity. Over recent
years, the collective investment by the three
existing operators in direct network infrastructure,
plus associated spectrum rights, has averaged
$400 million per year. On top of this, they have
invested hundreds of millions more dollars in core
technology systems to improve the delivery of
services across their respective wireless networks.
The Government is subsidising the rollout of
wireless networks into rural New Zealand where
sparse populations and challenging terrain
mean services cannot be provided on a normal
commercial basis. Overall, however, the wireless
sector has been a net contributor of revenue for
the Government:
• In the decade 2012 to 2022 (based on
investment to date and already-announced
commitments) the Government will contribute
almost $400m towards improved wireless
networks as part of the Rural Broadband
Initiative (RBI).
• During the same period, the Government has
already raised a total of $420 million from the
industry from auctions of 3G and 4G wireless
spectrum rights (before future revenue
from any 5G spectrum auctions is taken into
account).
• Moreover, about $250m of the
Government’s RBI funding is coming from
the Telecommunications Development
Levy, through which the broader industry
collectively pays a $50m annual levy (each
provider is levied according to its share of
total market revenue, with Spark’s share
almost $20m annually).
• By comparison, over the same 2012-2022
period, the Government will contribute
approximately $2 billion of taxpayer funding
towards improved fixed network services,
through the UFB and RBI programmes.
• Both the fixed and wireless network
deployments have improved coverage and
services for New Zealanders, but it is the
competitive environment of wireless networks
that has meant that those benefits have been
delivered largely without the need for a
taxpayer subsidy.
5G. The evolution towards a revolution.27

All other countries that are shaping up as early
adopters of 5G are also using a competition-
based industry structure to generate competition
and innovation.
Spark believes it is essential that wireless
operators have end-to-end control of their
networks so they can deliver services in new ways
including network slicing.
The specifications of 5G have been designed
to provide quite distinct and often conflicting
service requirements, so the way to cope with
this is to slice functions of the network tailored to
the performance characteristics of that service.
For example, massive IoT connections will require
centralised processing but the low latency
required for AR / VR will required processing
closer to the customer. Network slicing is defined
as an end-to-end capability – meaning that to truly
guarantee the service level of a network slice,
the operator must be able to stitch together the
relevant components across RAN, Transport and
Network Core.
Competition does not mean unnecessary
infrastructure proliferation. It means demand-
driven expansion to meet customer needs and to
maximise the network capabilities, and to create
innovation and client-specific services that will
get the best from 5G.
In practical terms, multiple competing networks
will not necessarily mean significantly more
infrastructure, such as cell towers, compared
with a single monopoly network, because
infrastructure is driven by the amount of data it
has to carry. Single towers having to carry more
traffic potentially means bigger towers to meet
the demand, which may present environmental
challenges especially in urban areas.
Spark considers the suggestion by Chorus that
the Government should contemplate a monopoly,
open-access network approach for 5G is founded
on incorrect assumptions about the way 5G
technology will show up for New Zealanders. As
discussed in this paper, 5G is best seen as an
overlay and advancement of existing wireless
technologies, not as a standalone network.
Any policy moves towards a monopoly radio
access network would create significant industry
uncertainty and would likely slow down the
introduction of 5G to New Zealand by years.
Chorus would need to build its own cell site
network (or arrange to access cell sites owned by
other operators) and either build its own network
core from scratch or force complex integration
with other wireless network operators. This would
result in the replication of network infrastructure
rather than leveraging the existing assets of New
Zealand wireless network operators - and have
the effect of reducing 5G network competition to
a common denominator.
This is not to say there aren’t opportunities
for Chorus to be involved with 5G. There is a
potentially attractive role for Chorus, as well as
the local fibre companies, to use their extensive
UFB fibre networks to provide high-quality
transport links from 5G cell sites and distributed
aggregation points in commercial arrangements
with wireless network operators.
5G. The evolution towards a revolution.28

Key Policy Considerations
In summary, New Zealand policymakers need to consider the following:
1. A competitive market model involving existing
wireless network operators and potentially
new entrants is the best way to ensure New
Zealand consumers and businesses benefit
from 5G.
2. A holistic view needs to be taken of the
environment within which 5G investments will
be made. Wireless networks are integrated
systems and they evolve over time
3. Innovative transport solutions will become
increasingly important as cell site counts
increase. New solutions such as the use of
shared-fibre networks and point-to-point
wireless may be required to make small-cell
sites economic. Regulatory support may be
required to obtain fair access to network
assets that are not provided directly by Spark.
4. Changes may be required to planning
practices to support the build of new
infrastructure and equipment, with the
development of a new planning standard
under the Resource Management Act.
5. C-band and mmWave spectrum work in
unison on a 5G network and need to be
considered in the same context and not as
separate functions.
6. C-band spectrum needs to be available in
sufficiently large blocks to operators so they
can build 5G networks with performance
exceeding what is possible today with 4G.
In practice this means at least 80MHz blocks
and ideally 100MHz blocks available to an
operator.
7. mmWave band spectrum should also be
released as soon as possible, as it is required
to realise the full benefits of 5G for New
Zealand consumers and businesses in tandem
with C-band deployment. To optimise data
speeds, spectrum blocks in the mmWave
band should be of at least 400MHz, with the
opportunity for an operator to acquire two
blocks totalling 800MHz.
8. The C-band and mmWave allocations should
be completed as soon as possible, to ensure
5G services can be delivered in time for the
2020-21 America’s Cup in Auckland as an
international showcase opportunity.
9. Regardless of operator, low-frequency
spectrum will be required to deliver the
benefits of 5G on a widespread basis into
rural areas (beyond small provincial towns).
The work to define 600MHz as a spectrum
band applicable to cellular networks should
continue at pace.
5G. The evolution towards a revolution.29

Beyond 5G
We think we’re well on the way to leading New Zealand into a 5G world.
But mobile technology evolution never stops and even before we’ve started deploying
commercial 5G networks, technologists, international standards bodies, and the wider mobile
industry, are starting to think about the next generation of mobile technologies again – 6G. So
what is 6G - or more accurately, what might it be?
More Data, More Spectrum
Just as 5G technology can deliver data faster and use larger spectrum blocks than 4G, 6G
will increase data speeds and spectrum block sizes again (provided that sufficient spectrum
is made available). So we expect to see an order of magnitude increase 6G relative to 5G -
meaning speeds in the tens of gigabits per second.
New Applications: ‘the Internet of Very Small Things’
Many of the applications being envisaged for 6G today tend to be similar to those already
identified for 5G – IoT, smart cities and transport systems, and automated vehicles or
equipment – but they are likely to be on a much grander scale. If 5G will bring us a world of
“connected things”, it is very possible that 6G will extend that concept to the micro scale. More
and more people are projecting that 6G will be the first generation of cellular technology to be
designed for nano devices and bio devices. Where 5G may connect lightbulbs in the home or
in the streetlight, 6G may connect the contact lenses we wear in our eyes.
Fully-integrated Artificial Intelligence (AI)
The possible scale of micro/nano technology, and the complexity of the networks that will
be required to connect that scale of devices, is extraordinary. Which means 6G networks
will need to be even more intelligent and autonomous than 5G networks. Near-ubiquitous,
instantaneous, connectivity, combined with artificial intelligence, is predicted to play a big part
in how 6G networks are managed and operated.
Conclusion
6G today is still a research topic. Much of the technology to enable it isn’t expected to become
a commercial reality for 10 years or more. And right now, it looks a lot like an extension of
5G. But as new technologies continue to emerge, especially in AI, materials, and nano and
bio technology, there is room for 6G to let us realise concepts and services that today exist
only in science fiction. A flexible, dynamic industry structure will be important to ensure New
Zealand’s future networks can adapt to and implement new technologies as they emerge.
5G. The evolution towards a revolution.30