Luke Coleman, CEO ATA Speech to Comms Connect

October 16, 2025-

Comms Connect, Melbourne

Space and land convergence — how rapid developments in satellite tech are revolutionising critical mobile connectivity

9:30am Thursday 16 October

Good morning everybody, and thank you to the event organisers for inviting me to speak today.

I’m Luke Coleman from the Australian Telecommunications Alliance, some of you may know the organisation from its previous title, Communications Alliance.

The organisation rebranded at the ATA a few months ago to make it 100% clear who we are and who we represent – our ambition is to be the clear and undiluted voice of the Australian telecommunications sector.

And that includes both mobile network operators and satellite service providers – and I am delighted to speak about what’s happening with both of those groups today, on the topic of “Space and land convergence — how rapid developments in satellite tech are revolutionising critical mobile connectivity.”

I’ve spent my entire professional career in the telecommunications industry, and I have never seen any technological development as exciting as Low Earth Orbit satellites, or LEOs.

In this industry we’ve become immune to hyperbole about “game changing” and “revolutionary” new tech, but this is one technology that well and truly lives up to the hype.

It is already shifting the foundations of the telecommunications industry in Australia and around the world – for the better.

So today I’ll talk about satellite and mobile convergence in three parts.

First, I’ll start with a very recent history of satellite technology, and how we’ve seen quantum leaps in satellite technology over the past ten years.

Second, I’ll turn to the state of play today: where the lines are already being blurred between satellite and terrestrial networks, devices, and services.

And finally, I’ll look at the implications for the market and for policymakers, as these new technologies solve policy problems that have existed for decades.

So let’s start with a very recent history of satellite technology. It is mind blowing how far we’ve come in just ten short years.

It was a decade ago almost to the day – the 1^st of October 2015 – that NBN launched the first of its two Sky Muster satellites.

I was the regional telecommunications policy adviser in the Communications Minister’s office, and it was quite a moment.

I’m still annoyed that I didn’t get to go to French Guiana to watch the rocket launch.

To illustrate how far we’ve come in just a decade, let me give you a reminder of what people in regional and remote Australia were stuck with before Sky Muster launched.

Prior to the launch of Sky Muster, there was a short-term policy fix called the NBN Interim Satellite Service.

It provided maximum speeds of 6 megabits per second, although typical users speeds were less than half that, and a monthly download quota of 3GB of data.

In contrast, the average monthly data allowance on metropolitan fixed-line broadband networks in 2015 was over 400GB – more than 100 times that amount.

The digital divide was more like a yawning chasm between the cities and the bush.

The Interim Satellite Service was a technical and a political headache.

Constant complaints about poor throughput, technical issues, and limited uptake – it left some real brand damage for satellite services among users in regional and remote areas.

A year after the launch of Sky Muster 1, on the 6^th of October 2016, Sky Muster 2 launched.

At the time, these satellites were world-class. They were capable of delivering 25 megabits per second peak download speed, five megabits per second peak upload speed, and markedly higher download capacities.

The two satellites provided coverage to all of Australia and our external territories including Norfolk Island, Christmas Island, Lord Howe Island, Macquarie Island, and Cocos Island.

Sitting in orbit 36,000km above the earth, they use 101 Ka-Band spot beams to cover all of Australia, and connected to ten ground stations.

But even with this new world-class satellite system in operation, the digital divide was still there.

In fact, the divide only grew – despite the leap in performance and data capacity that these new satellites offered, the rapid rollout of the NBN in the cities meant that metro users were getting far faster internet speeds and far higher download capacities.

While the bush was getting better services than ever before, the digital divide continued to grow.

These state-of-the-art satellites solved one policy problem – they delivered a step-change in broadband quality for remote areas – but they didn’t solve every policy problem.

As the NBN rollout was progressing at pace, the Government introduced new laws to make NBN the default voice and broadband provider, called the ‘Statutory Infrastructure Provider’.

But – NBN’s satellite services were partially excluded because satellites weren’t capable of providing voice services good enough to replace Telstra’s fixed-line copper network.

It all came down to the laws of physics, and in technical terms, Round-Trip Time – that is, the amount of time it takes for a signal to get from earth to a satellite and back again.

Geostationary satellites like Sky Muster sit around 36,000km above the earth, so it takes just under 300ms for a signal to get from earth to the satellite and back again.

When the signal needs to do that trip twice to transmit a call back and forth, that’s 600ms for a double-hop back and forth to the satellite – far too much latency for reliable voice calls.

ACCC measurements found an average latency on Sky Muster of 664 milliseconds[1] – fine for most internet users, but not a viable replacement for fixed-line voice services.

The International Telecoms Union (ITU) recommends that one-way latency should not exceed 150ms for voice calls, meaning a maximum 300ms return trip – physically impossible for a satellite 36,000km above the earth to do.

So we wound up in a situation where the Commonwealth Government had invested around $2 billion[2] in the satellite program overall, but had to continue paying Telstra to operate its fixed-line copper network across the entire satellite footprint to provide voice services.

I’ll discuss the policy implications of this a little later on.

Of course, NBN’s satellites were just two of many competitive satellites providing services in Australia – alongside Optus’ fleet, and a range of international operators.

But they all faced the same problem – you can’t break the laws of physics.

So with that very brief recent history of satellite technologies out of the way, we’ve reached the part of the story where things get really interesting.

Because while the laws of physics can’t be broken, they can be outsmarted by giant leaps in technology.

Just over two years following the launch of Sky Muster 2, and after three years of planning, a company named Starlink launched its two Low Earth Orbit test satellites in February 2018.

Fifteen months later in May 2019, Starlink launched its first operational LEOs – 60 of them.

In 2020 Starlink launched a public beta program to test the system, before going commercial in 2021 offering broadband speeds of around 150Mbps.

By 2022 Starlink had launched a business-grade product, offering speeds of up to 500Mbps on using high-performance antennas.

Around the same time it also launched its Roam product, which allowed for portable antennas that could be used on-the-go in caravans and boats.

And by 2023, Starlink began deploying its second-generation of satellites, the ‘v2 Mini’, which were around three times larger and delivered around four times the capacity of its previous generation.

But while they might be the most recognisable name in the LEO space, Starlink is not the only game in town.

AST SpaceMobile, Iridium, Omnispace, SES/Intelsat, OneWeb, TeleSat, and Amazon Kuiper are all in the game, and competition in the space is heating up.

Amazon recently signed a contract with NBN Co to use Kuiper to replace the Sky Muster satellites.

Amazon’s constellation is planned to include more than 3,200 LEO satellites, which began deploying in April with its first operational launch.

There are currently 78 Kuiper satellites in orbit, after three successful launches in less than three months, with many more to come.

So what’s the magic behind LEO satellites that makes them such a giant technological leap from GEO satellites?

Here’s the slide I showed earlier, where traditional satellites are limited by their orbit 36,000km above the earth.

Being that high has its advantages – at such a high altitude, you can cover all of Australia’s landmass and our external territories with just two satellites.

Think of it like shining a torch beam at the ground – the higher you lift the torch, the wider the radius of the beam.

But it also has disadvantages – the further away the torch, the weaker the light.

And so it is with data speeds and latency.

LEO satellites sit between 300 – 500km above the earth’s surface – dramatically closer than GEO sats.

As a result, you need a lot more of them to provide coverage. Those torch beams are much smaller, and it would take dozens, or even hundreds, of LEO satellites to provide the same coverage of earth as just a couple of GEO satellites.

And that’s exactly what they’re doing.

Today, there are around 8,000 Starlink satellites in orbit – that’s around two-thirds of the total number of satellites orbiting earth, operated by one company.

The company has filed applications to launch as many as 42,000 LEO satellites in total, so they’re only at around 20% of their total planned constellation.

Each dot you see on that map of Australia is a Starlink satellite.

And unlike geostationary satellites that sit comfortably in their spot, shining down that beam, LEOs are constantly spinning around the earth at tens of thousands of kilometres per hour.

For illustrative purposes on this next slide, we’re seeing the coverage of the Starlink satellites just over NSW and south-east Queensland – the bright orange lines show their orbits, and as you can, when they’re working as a group they provide contiguous coverage.

If we lit up every one of those dots, all of Australia would be shown to have coverage.

Now here’s where we start to blur the lines between space and land, between terrestrial mobile services and satellite broadband services.

More than 99% of Australians homes and business have mobile coverage from at least one operator – Telstra. More than 98% of premises also have coverage from Optus and TPG, under their network sharing arrangements.

But while around 99% of premises have coverage, that’s still only around a third of Australia’s total landmass.

And the reality is, that other two-thirds of Australia’s landmass is never going to have mobile coverage. There is just no economic case to build more mobile towers in very remote areas.

Even successful Government co-contribution programs like the Mobile Black Spot Program have run out of steam, to the point where mobile network operators aren’t willing to co-invest nearly as much as they once were.

But when you add LEOs, you instantly have 100% coverage of all of Australia.

A Starlink dish, connected to a Wi-Fi router, anywhere in Australia, will give you voice and broadband speeds that are equivalent to being on the NBN in a major city.

All Australian mobile network operators support Wi-Fi calling, so if you have any standard mobile phone connected to a Starlink Wi-Fi network, you’ll be able to make a mobile phone call with voice quality that is imperceptible compared to a normal mobile or even fixed-line voice call.

And with the ability to use Starlink’s Roam service, you can take your dish anywhere and effectively set up a micro-mobile network anywhere in Australia.

But that’s nothing new – you’ve been able to do that for years.

Where it gets really interesting is when you don’t even need the Starlink dish or the Wi-Fi router, and you can just connect directly with your phone.

In January 2024, Starlink launched to the first of its satellites with direct-to-device capability.[3]

Starlink, being American, calls this direct-to-cell, and others call it direct-to-handset, or satellite-to-mobile, so we still need to take a vote on what acronym we’re going to use.

I’m going with direct-to-device.

The real beauty of direct-to-device is that unlike traditional GEO satellite phones, which were typically quite bulky with large antennas, this works with standard mobile devices like iPhones and Samsungs.

And the secret sauce is spectrum[4].

And this is where things get really interesting.

Because there are two different types of direct-to-device spectrum, which won’t only dictate the types of services available, but will also set the stage for a power struggle between mobile network operators and satellite operators over control of the networks in the future.

Think of a bit like VHS vs Betamax in the 1980s, if you’re old enough to remember.

First, there’s IMT spectrum – which stands for International Mobile Telecommunications.

That’s the terrestrial spectrum used by Telstra, Optus, and TPG to deliver 4G and 5G services from their mobile towers.

Those spectrum bands are harmonised around the world for mobile services.

And it’s the spectrum that the chipsets in all standard mobile phones recognise and connect to.

So no matter where you go around the world, chances are you’ll be able to connect to the local 4G or 5G network with the device you bought in Australia.

Direct-to-device services using IMT spectrum can be operated under the current spectrum licensing framework without the need for explicit approval from Australia’s spectrum regulator, ACMA.[5]

And this is what Telstra is using in its partnership with Starlink to enable SMS messaging outside of its network coverage[6], where Telstra allows Starlink to utilise its IMT spectrum to connect to its satellites.

Live trials of text messaging via Starlink commenced in April, with more than 55,000 messages tested in the first month, as well as sending GPS coordinates, and most importantly, emojis.

Optus has also announced a partnership with Starlink for direct to device[7].

Starlink completed its first-generation constellation of 600 direct-to-device satellites in June this year[8].

Second, there’s MSS spectrum, which stands for Mobile-Satellite Services.

MSS spectrum requires your device to have a chipset that recognises MSS frequency bands, and effectively makes your mobile device act like a tiny satellite earth station that can fit in your pocket.

These chips are in more recent devices – Apple iPhone 14 and above, and newer Google Pixel 9 and Samsung Galaxy 25 models.

It’s this band that enables Apple, Google’s proprietary satellite services like messages, location sharing, emergency SOS, and roadside assistance.

And unlike IMT spectrum used by mobile network operators, MSS spectrum is only available to satellite operators.

The fight for control over direct to device connectivity will largely hinge on who controls the spectrum. There’s only so much of it to go around.

Satellite services on iPhones are enabled by Apple’s agreement with satellite service provider Globalstar, utilising its MSS spectrum[9]. Similarly, Google is working with Skylo[10].

It’s also this spectrum band which Starlink is using to deploy its second-generation direct-to-device satellites, having paid US$17 billion to purchase MSS spectrum from EchoStar just last month.[11]

These second-gen satellites will sit just 360km above the earth[12] to optimize the link between the devices and satellites.

So we’ve got two competing technologies to deliver direct to device connectivity, and there’s nothing stopping both working simultaneously.

One device could be connected to one satellite operator via MSS spectrum for certain services, and to a different satellite operator for other services, and it could even simultaneously be connected to a terrestrial mobile network.

If you had dual-SIMs in that device, you could be connected to two terrestrial mobile networks.

And this is just the beginning of space and land convergence.

Currently, these direct to device services are relatively low-bandwidth – text messages and location sharing during emergencies.

These services are being marketed as emergency services – they’re not being positioned as substitutes for mobile coverage.

But when the next generation of MSS band LEOs satellites is operational, expect to see voice calls and higher-bandwidth data services like video calls.

Within a few short years, expect to see MSS chips as standard in the majority of smartphones – satellite connectivity will be seamlessly integrated with terrestrial mobile networks.

100% universal, global mobile coverage is likely to be a reality by the end of this decade. Arguably it’s already a reality today, albeit in a limited form.

And that thought brings me to the third and final part of my remarks today: how the convergence and space and terrestrial connectivity is changing the market, and solving policy problems that have existed for decades.

The sudden arrival high-speed internet anywhere in Australia has seismic implications for the market.

Let me give you a few examples.

For businesses operating in remote areas, access to high-speed, high-capacity broadband has historically been something that’d cost tens of millions of dollars.

In Australia, the mining and resources industry is as dependent on connectivity just as much as on trucks and diggers.

To get connectivity on a remote mining site, your options were either building extremely expensive fibre-optic links, which would probably cost tens of millions of dollars to deliver, or microwave connection which were lower-bandwidth and still expensive – building towers in remote areas isn’t a cheap exercise.

The arrival of LEO services completely upends the market.

What once cost tens of million of dollars can now be delivered for tens of thousands of dollars.

Here’s a picture of a mining site in remote South Australia, operated by Heathgate Resources, that has bonded multiple LEO dishes.

The result? They went from having 2Mbps of data capacity via microwave and GEO satellite, to 1.3Gbps downlink and 250Mbps uplink[13] via LEOs.

4 years ago, an equivalent amount of capacity would have required a fibre connection costing tens of millions of dollars. It’s completely changed the market.

And of course this has seismic implication for Government funding programs as well.

All of Australia’s regional telecoms funding programs have been built on the premise that regional telecom networks are inherently loss-making, and therefore will always require subsidies to operate.

It’s why we have the Universal Service Obligation, a $250 million annual payment to Telstra to keep its old copper network alive.

It’s why we have the regional broadband scheme, a $1 billion annual cross-subsidy to fund NBN Co’s regional satellite and fixed-wireless networks.

But when anyone can buy LEO dish for a few hundred dollars, the underlying premise for these policy interventions no longer makes sense.

It’s the same case for funding programs to increase mobile coverage.

And it’s why the Government has announced legislation to establish a Universal Outdoor Mobile Obligation or UOMO on mobile operators, to require them to provide direct to device services as the market develops.

The digital divide that has plagued policymakers for decades has disappeared virtually overnight.

When you can access metro-comparable broadband speeds and mobile connectivity anywhere in Australia, it’s a question if there even is a digital divide anymore.

And of course, for this audience, the implications for emergency services and disaster recovery are incredible.

When power outages take down terrestrial mobile networks during natural disasters, being able to fall back on LEO connectivity and portable generators can literally be a life-saver.

High-bandwidth coverage anywhere in Australia means emergency service operators can almost instantly deploy local Wi-Fi networks, or even small-scale mobile networks using LEOs for backhaul connectivity.

When direct-to-device technologies have advanced over coming years, that won’t even be required as devices will connect directly to LEOs.

Space and land technologies have already converged. The technology is advancing at a rapid pace.

And by the end of this decade, it’ll be hard to even tell the difference anymore.

So I’ll conclude with a brief recap of my remarks today.

First, we’ve seen quantum leaps in satellite technology over the past ten years, where typical services have gone from 2Mbps to 200Mbps.

Second, we looked at the state of play today: where the lines are already being blurred between satellite and terrestrial networks, devices, and services.

And third, we considered the implications for the market and for policymakers, as these new technologies have solved policy problems that have existed for decades – virtually overnight.

Thank you.

END

[1] https://www.accc.gov.au/media-release/broadband-performance-of-satellite-services-measured-for-the-first-time

[2] https://www.smh.com.au/politics/federal/sky-muster-nbns-fancy-satellite-could-explode-is-as-big-as-an-elephant-and-was-initially-unloved-20150930-gjxwx8.html

[3] https://www.starlink.com/public-files/DIRECT_TO_CELL_FIRST_TEXT_UPDATE.pdf

[4] https://www.acma.gov.au/sites/default/files/2025-05/Regulatory%20guide_Operation%20of%20an%20IMT%20satellite%20direct-to-mobile%20service_0.pdf

[5] https://www.acma.gov.au/sites/default/files/2025-05/Regulatory%20guide_Operation%20of%20an%20IMT%20satellite%20direct-to-mobile%20service_0.pdf

[6] https://www.telstra.com.au/exchange/telstra-satellite-to-mobile-connectivity–our-latest-trials-and-

[7] https://www.optus.com.au/about/media-centre/media-releases/2023/07/together-optus-and-spacex-plan-to-cover-100-percent-of-australia

[8] https://www.spacex.com/updates#dtc-gen2-spectrum

[9] https://www.lightreading.com/satellite/what-is-globalstar-building-for-apple-

[10] https://www.skylo.tech/newsroom/google-and-skylo-expand-satellite-connectivity-to-pixel-10-series-and-unveil-pixel-watch-4#:~:text=On%20the%20Pixel%2010%20Series,to%20vehicle%20accidents%20on%20highways.