Today's iPhones and Androids cannot communicate directly with satellites in space. Instead, they must connect to nearby cell towers on the ground.

But that may soon change.

A large range of companies – from tiny startups to some of the biggest corporations on Earth – are racing each other to connect regular smartphones directly to satellites. According to analyst Lluc Palerm of Northern Sky Research (NSR), this represents the largest opportunity in the history of the satellite communications industry, worth $66.8 billion over the next ten years. That's almost double what should be generated by existing, standard satellite services over that period.

However, the phone-to-satellite market is still in its infancy. It's also weighed down by significant technical challenges. But it's the regulatory and business hurdles involved in mashing together the cellular and satellite industries that could be hardest to overcome.

"There's a lot of complicated pieces," agreed AST SpaceMobile Chief Strategy Officer Scott Wisniewski. AST SpaceMobile is one of the many companies playing in the sector.

Nonetheless, there is a distinct chance that, within the next 12-18 months, a large number of Americans will be able to send text messages from virtually anywhere in the US, from the jagged peaks of Colorado's Rocky Mountains to the barren Badlands of South Dakota, as long as they have a an unobstructed view of the sky.

And, if rumors of the upcoming iPhone 14 are true, some Americans may be able to do that in just a few weeks.

Two paths to space

As noted by NSR's Palerm, there are two basic ways that smartphones may connect to satellites. They're either going to use spectrum intended for satellites, or spectrum intended for smartphones.

Elon Musk's SpaceX, through its Starlink subsidiary, appears to be pursuing both paths simultaneously.

During his media event with T-Mobile's Mike Sievert last week, Musk said Starlink plans to build satellites that will support transmissions in T-Mobile's PCS spectrum. According to PhoneScoop, those transmissions will work in the G Block of PCS Band 25, which runs across 1910-1915MHz for uplinks and 1990-1995MHz for downlinks. That's a licensed spectrum band intended for cellular communications, and it's already supported in virtually all of T-Mobile's existing phones.

Musk explained that Starlink's planned second-generation satellites will be able to provide T-Mobile with 2-4Mbit/s capacity per sector. Starlink's current sectors are about 15 square miles. Thus, T-Mobile initially only plans to support text messaging instead of high-bandwidth applications like voice calls and streaming video.

But that still could represent a huge opportunity if other network operators join T-Mobile in signing up for Starlink's planned services. "What if they cover the globe with 20 different operator partnerships, offering text messages and limited email only? They could reach a billion customers and pull in ARPU [average revenues per user] of $1 per customer per month. A billion bucks per month? That's a winner," wrote analyst Joe Madden, with Mobile Experts, in a recent FierceWireless article.

However, Starlink's plans with T-Mobile still require FCC approval. After all, the agency regulates traffic over airwaves in the US, and it would need to modify its rules to allow transmissions in the PCS G Block to go to space and not just nearby cell towers.

Starlink hopes to begin providing initial services for T-Mobile customers by the end of 2023.

But that's not Starlink's only gambit. The company is also asking for FCC permission to use a slice of the 2GHz spectrum band to "augment" its satellite services. Specifically, according to PCMag, the company hinted at "next-generation services for mobile users" stemming from its acquisition of Swarm Technologies last year.

Interestingly, Starlink's pursuit of 2GHz spectrum puts the company again into direct competition with Dish Network, which already has permission from the FCC to use 2GHz – and a variety of other spectrum bands – for its nascent 5G network. According to Starlink, Dish isn't using the spectrum and therefore shouldn't have access to it. "There is scant evidence that Dish is actually providing MSS [mobile satellite service] service to anyone, anywhere," Starlink argued.

Dish quickly slapped back at Starlink's request. "SpaceX, a company led by the richest person in the world, has already received authorization to use 15,000MHz of spectrum – free of charge – from the FCC," a Dish representative told SpaceNews.

(The bickering between Starlink and Dish comes as little surprise. The companies are also locked in a heated battle over how best to use the 12GHz band.)

But there's one big hitch in Starlink's apparent plans to use the 2GHz band to connect to smartphones (presumably in international locations not covered by the company's new deal with T-Mobile). Customers' phones must also support transmissions in the 2GHz band, and most phones do not yet do that.

"A mainstream 5G device will have the potential to directly communicate with a satellite if chipsets incorporate this capability," explained Palerm, of NSR. "So, the main challenge for these traditional satellite operators with spectrum rights across the globe is to convince Qualcomm, MediaTek and the other 5G chipset manufacturers to incorporate those capabilities (spectrum, waveform modification, etc.) into their future designs."

Enter the iPhone 14 Pro

That's exactly what Globalstar appears to have done. The company is widely expected to power a phone-to-satellite service for Apple's next iPhone, scheduled to be unveiled next week.

Globalstar, a longtime satellite operator, inked an agreement with a mystery customer in 2020 that has so far generated tens of millions of dollars in revenue for the company. Most recently, Globalstar announced that its mystery customer will spend roughly $310 million to purchase 17 new Globalstar satellites. According to the B. Riley Securities analysts, speculation about the identity of that customer has ranged "from Dish to Google, and now Apple."

More importantly, Globalstar has been working to develop an ecosystem around its 2.4GHz S-band spectrum, dubbed Band 53. The spectrum traces its origins to satellites, but it can be used by mobile devices. Already Globalstar has gotten Qualcomm to support Band 53 in its X65 chipset for smartphones. Apple famously switched in 2019 from Intel to Qualcomm chips for its new iPhones. That's about the same time Apple's Tim Cook reportedly became personally interested in a research project that would use satellites to bypass terrestrial wireless networks.

Analyst Tim Farrar with TMF Associates recently suggested that Apple would begin offering a free messaging service using Globalstar's spectrum as soon as its new iPhones go on sale. However, it's likely that Apple's phone-to-satellite service will be exclusive to its most expensive "Pro" iPhone models, at least initially.

If that's the case, Apple's initial satellite-usage payments to Globalstar might be relatively small considering iPhone Pros account for just a portion of the vendor's overall iPhone sales.

Globalstar isn't the only satellite company toting spectrum that could be used by mobile devices. For example, satellite operator Iridium recently announced that it "entered into a development agreement to enable Iridium's technology in smartphones." Separately, Omnispace is working with a mobile network operator in the Philippines on a phone-to-satellite service. And billionaire Charlie Ergen has hinted at "cross-pollination" between his EchoStar satellite business and the terrestrial 5G network he's building through Dish Network.

But such efforts will still require phone makers to add support for Iridium, Omnispace and EchoStar satellite spectrum into their phones.

Cell towers in the sky

Two startups, AST SpaceMobile and Lynk, hope to neatly bypass this problem. Like Starlink's deal with T-Mobile, they're planning to operate satellites that support transmissions in existing cellular spectrum bands. That way, they can immediately begin supplying connections to billions of existing smartphone users around the world.

Right now though, it's a race to see which company can get the approvals, deals and satellites necessary to launch such services.

For its part, AST SpaceMobile hopes to launch its second prototype satellite, confusingly dubbed BlueWalker 3, early next month. The company doesn't expect to begin launching satellites for its commercial service until the end of 2023.

But AST SpaceMobile already has carrier agreements with the likes of Vodafone, Rakuten and Orange. According to Wisniewski, the company's strategy chief, AST SpaceMobile also has a commercial agreement with AT&T.

Figure 1: AST SpaceMobile plans to assemble dozens of satellites at its Midland, Texas, headquarters.
(Source: AST SpaceMobile. Used with permission.)

But AT&T isn't yet ready to begin talking about phone-to-satellite services. "There's a market out there to keep people connected all the time," AT&T CEO John Stankey told Axios recently. But he declined to share details about any plans from AT&T. "I think we'll see the market develop where there's a variety of different alternatives and solutions," he said.

AST SpaceMobile hasn't yet disclosed exactly which cellular spectrum bands it might support, but they could range anywhere from 700-960MHz and 1600-2100MHz.

Interestingly, AST SpaceMobile has raised almost $600 million so far, but it has burned through almost $400 million of that. The company counts 300 full-time employees.

Meanwhile, AST SpaceMobile rival Lynk appears to be doing more with less. The company has raised a total of just $30 million and counts 40 employees. But Lynk has already launched five different test satellites. The company's sixth satellite – dubbed Lynk Tower 1 – launched in April and can work across the cellular 600-960MHz bands. That's the satellite that Lynk expects to use to begin offering commercial services by the end of this year.

"With small teams you can do amazing things," said CEO Charles Miller.

Lynk has announced commercial deals with a number of international wireless network operators, but none in the US. "We are testing with multiple big US operators," Miller said, adding: "We have agreements with at least one big US operator."

But he declined to provide details. He also declined to discuss the regulatory approvals Lynk may need to offer commercial services in the US market.

Regardless, Lynk hopes to launch three more commercial satellites by December. Miller said that ought to allow the company to begin offering coverage in locations across the globe a few times every day. That's because phones will only be able to connect to Lynk's satellites as they pass overhead. Still, Miller said, that ought to be good enough for some emergency and IoT-style services.

Bigger and better

Although a number of companies have promised to connect phones to satellites – and standards along those lines are in development – the technical challenges involved in those kinds of communications are substantial. After all, most traditional satellite services rely on giant satellites connecting to bulky satellite phones with big antennas. Further, most standard smartphones are designed to transmit their signals just a few miles, to a nearby cell tower. For satellites to pick up those signals from hundreds of miles away – while traveling thousands of miles an hour – is an engineering feat just short of a miracle.

"This is really quite a difficult technical challenge," Musk agreed during his media event with T-Mobile.

As a result, most satellites designed to connect to regular phones are much bigger than the current low-Earth orbit (LEO) satellites operated by companies like Starlink and OneWeb. That extra size is necessary to help pick up customers' smartphone transmissions, and to weed out transmissions from other nearby phones.

For example, at 269 square feet, the second-generation Starlink satellites Musk hopes to use for T-Mobile's service are roughly five times the size of Starlink's current satellites. And AST SpaceMobile's planned BlueWalker 3 test satellite clocks in at 693 square feet. "Size matters!" proclaimed Abel Avellan, AST SpaceMobile CEO, in a LinkedIn post.

Lynk, meantime, wouldn't provide the exact size of its current Tower 1 satellite, noting only that the company's future satellites will be roughly the same size as Starlink's second-generation satellites.

Another factor is the number of satellites supporting the service. LEO satellite services in general are designed to work through a constellation of satellites spinning around the Earth, rather than one geosynchronous satellite that stays in the same place above the Earth's surface. One Lynk satellite would only provide connections in the US twice a day, for example. That's why Lynk hopes to launch 50 satellites by the end of 2023.

Similarly, AST SpaceMobile plans to launch five satellites during 2023, and its eventual goal is to operate 168 satellites in total.

Starlink currently counts more than 2,000 LEO satellites in operation, giving the company the network capacity to provide 90 Mbit/s services to roughly 500,000 customers who have purchased the company's $500, stationary, pizza-sized receivers. While that's not good enough for the FCC, considering the agency rejected Starlink's request for government subsidies, it's good enough for Royal Caribbean Cruises to use Starlink services on all of its cruise ships.

However, there's one final element affecting the operation of a satellite constellation: getting that constellation into orbit. Due to the size of Starlink's second-generation satellites, the company plans to use SpaceX's Starship heavy-lift reusable rocket for such launches. SpaceX's Starship is the tallest and most powerful launch vehicle ever built, but it is still in the testing phase.

AST SpaceMobile and Lynk plan to use SpaceX's existing Falcon 9 launch rockets to get their own satellites into space. SpaceX has emerged as a critical space transport option for US companies following a broad shift away from Russian rockets due to Russia's unprovoked war on Ukraine.

Lynk's Miller said he's not concerned about using SpaceX rockets to launch Lynk satellites despite the fact that those satellites will compete directly with the second-generation satellites from SpaceX's Starlink subsidiary.

Finally, such satellites also pose regulatory concerns. Lots of big LEO satellites could further crowd Earth's orbit, raising concerns ranging from space junk to astronomy obstructions. Thus, it's not clear whether regulators will approve Starlink's second-generation satellites, much less the proposed constellations from the likes of AST SpaceMobile and Lynk.

Operators, roaming partners or competitors?

The rise of satellite options for smartphone users also raises one major, final question: What does this all mean for AT&T, T-Mobile and other terrestrial mobile network operators?

Lynk and AST SpaceMobile have been very clear in their intentions: They want to operate just like a roaming partner for terrestrial mobile network operators (MNOs). "We will never do business without an MNO," said Lynk's Miller. "We will never go direct-to-consumer."

AST SpaceMobile laid out its entire business plan in 2020 as it prepared for its initial public offering (IPO). The company hopes to send texts to customers who travel outside of their operator's terrestrial coverage area, asking them if they want to activate a "SpaceMobile Day Pass." It would then share the resulting revenues with its mobile operator partners; the company already has a 50/50 revenue share agreement with Vodafone.

Starlink might be different though. Musk said the company is open to agreements with other mobile network operators beyond T-Mobile, but that might be complicated by its plans to support T-Mobile's 1900MHz PCS spectrum. That band isn't widely used across Europe and other countries. That, according to Lynk's Miller, could indicate that Starlink may also pursue a direct-to-consumer strategy by using the 2GHz band.

Globalstar and Apple certainly would appear to be planning a direct-to-consumer strategy. Such an offering would circumvent Apple's traditional wireless network operator partners thanks to Globalstar's own licensed spectrum holdings. That could exacerbate tensions between Apple and its wireless carrier partners.

T-Mobile, meanwhile, has provided a look at how terrestrial network operators might pursue the topic. T-Mobile plans to add satellite messaging capabilities into its most expensive plans for free, and may charge other customers a per-month fee for the service. Thus, T-Mobile's deal with Starlink might be similar to T-Mobile's deal with Netflix, whereby T-Mobile offers Netflix services for free on its most expensive plans. That broadly fits into T-Mobile's overall strategy of encouraging customers to upgrade to its most expensive plans by packing those plans with everything from unfettered 5G to streaming video to satellite messaging.

Broadly, this all may just be the beginning of a broader merging between the cellular and satellite industries. For example, Verizon plans to use Amazon's forthcoming Project Kuiper LEO satellites to connect its rural cell towers to the Internet. Meanwhile, AT&T is planning a similar setup with OneWeb's own growing constellation of LEO satellites. It's likely that both OneWeb and Project Kuiper are considering adding cellular capabilities to their LEO satellites in the same way Starlink would with its second-generation satellites.

Companies with satellite spectrum – such as Iridium – could also follow Globalstar's lead by partnering with device makers like Samsung.

And hanging over this whole sector are specifications under discussion at the 3GPP that would add non-terrestrial networks (NTN) directly into the 5G standard. Indeed, NTN capability is already scheduled to be part of the 3GPP's Release 17 package of specifications, to be approved later this year.

In the meantime, executives like Mike Sievert and Elon Musk are clearly working to stake out their positions in the "final frontier."

Related posts:

— Mike Dano, Editorial Director, 5G & Mobile Strategies, Light Reading | @mikeddano