How One Week Turned LEO, GEO and D2D into Europe’s Messiest Love Triangle
LEO/NGSO went from PowerPoint to policy and product, GEO reminded everyone it still pays the mortgage, D2D crossed the line from hype to infrastructure, and IRIS² graduated from Brussels buzzword to funded multi-orbit reality. Let’s unpack the highlights, with an appropriate mix of analysis and eye-rolling.
Starlink’s Direct-to-Cell goes live in Europe: Kyivstar as the first real test case
The headline shift for the week is simple: direct-to-device in Europe is no longer a future capability; it’s a live network feature that ordinary customers can use, at least in one of the hardest operating environments on the continent.
On 24 November 2025, VEON announced that Kyivstar has launched Starlink Direct to Cell service across Ukraine, making it the first mobile operator in Europe to offer Starlink’s satellite-to-phone connectivity. The press release is very explicit: this is real commercial service, on regular 4G smartphones, with coverage “across the entire territory of Ukraine” except occupied areas, border regions, and active combat zones. The initial phase is SMS only, with plans to add voice and data in 2026.
What makes this different from yet another “world-first satellite SMS” press event is the context and the scale. Kyivstar serves more than 22.5 million mobile customers, and the service is explicitly framed as a resilience tool: connectivity during prolonged blackouts, in recently de-occupied territories, and in areas where terrestrial networks are damaged or under restoration. This is a national-scale test of whether LEO D2D can function as a de facto civil defence backbone.
From a technical and commercial standpoint, this is also the first visible proof of Starlink’s claim that its Direct to Cell constellation can behave like a roaming partner, abstracted behind operators’ existing core networks. The VEON release describes Direct to Cell as “acting like a cell phone tower in space,” integrated through standard roaming-style mechanisms, which is exactly the narrative SpaceX has been pushing to nervous MNOs for the last two years. The war in Ukraine has effectively become the high-stakes beta environment where that architecture gets tested at scale.
Strategically, Kyivstar’s launch also validates the “MNO as hero” storyline Starlink and others have been selling regulators who are rightly worried about satellite players bypassing local operators. By having the Ukrainian Ministry of Digital Transformation front-and-center in the announcement, and stressing that connectivity is free to Kyivstar users at no extra cost, VEON and Starlink are presenting D2D as a national infrastructure upgrade rather than an over-the-top threat. It’s a clever move: in a war-time setting, arguing against redundant satellite coverage is politically suicidal, and that optics will not be lost on other European regulators watching this unfold.
If this were a movie, this would be the scene where the beleaguered kingdom gets a shipment of game-changing weapons from an eccentric ally. Everyone cheers, but the audience knows the real drama will be about dependency, control, and who sets the rules when the dust settles.
Proximus Global and Starlink: BICS steps up as Europe’s D2D gatekeeper
If Kyivstar is the first European operator to press the big red “ON” button for Starlink Direct to Cell, Proximus Global spent this week making sure that when the rest of Europe inevitably follows, a Belgian carrier-services business is sitting in the tollbooth.
On 25 November 2025, Proximus Global announced that its wholesale arm BICS has become Starlink’s preferred IPX provider in Europe for Direct-to-Cell services. The official press release spells it out: Starlink will use Proximus Global’s IPX network as the bridge to connect mobile network operators to its D2D constellation, both for today’s SMS-oriented services and for a future “next-generation network that will be capable of providing broadband service to smartphones.”
IPX has always been the quiet infrastructure behind roaming and interconnect; now it’s quietly becoming the control plane for non-terrestrial networks. Proximus Global emphasises that its IPX behaves like a roaming exchange, presenting Starlink as just another roaming partner while hiding the messy details of beam-hopping constellations, spot-beam handovers, and spectrum gymnastics from the MNOs. It is, in other words, an elegant way to let European operators dip their toes into satellite-to-phone without having to re-architect their core networks or negotiate directly with Elon’s team of Very Online Salesmen.
The subtle but important twist is that Kyivstar is called out as the first European operator to benefit from this partnership in the same Proximus Global release. The press note explicitly says that Kyivstar will be the first launch customer on this Starlink–BICS arrangement, using the IPX fabric to enhance resilience “particularly in emergencies when network infrastructure is harmed.” So within a 48-hour window, we saw both ends of the architecture click into place: VEON announcing live Ukraine service, and Proximus/BICS quietly positioning itself as the preferred intermediary for everyone else who wants in.
For the satellite industry, the broader implication is that D2D’s commercial power may end up concentrated in the middle layer, not just at the constellation operator. If enough operators standardise on BICS (or similar IPX intermediaries), the entity that controls routing, QoS, and settlement across terrestrial and non-terrestrial domains becomes the real strategic chokepoint. In gaming terms, Starlink is farming LEO kills, but Proximus is the one controlling the loot drop.
The messaging from Proximus Global is also notably soothing to terrestrial operators: the CRO goes out of his way to say that satellites are “not inherently disruptive” and that “there’s room for everyone.” It’s the rhetorical equivalent of a dragon explaining that it mostly eats livestock and only occasionally knights, so really, what’s the problem? The reality, of course, is that once customers have experienced “coverage wherever you can see the sky,” tolerance for patchy terrestrial coverage will drop, and MNOs that don’t plug into these IPX-mediated satellite fabrics will be asked some uncomfortable questions by their enterprise clients.
Amazon Leo’s Ultra antenna: the enterprise LEO gauntlet is thrown
While SpaceX was busy turning European MNOs into roaming partners, Amazon chose this week to remind everyone that its own LEO constellation is not content to be a perpetual science fair project.
In a 24 November 2025 article on its corporate site, Amazon detailed the first real product incarnation of its rebranded constellation, now called Amazon Leo. The piece introduces Leo Ultra, a full-duplex phased array terminal that Amazon cheerfully describes as “the fastest commercial phased array antenna in production”, supporting download speeds up to 1 Gbps and upload speeds up to 400 Mbps. The article also announces an enterprise preview program so selected customers can begin testing the network ahead of broader rollout in 2026.
Unlike Starlink’s consumer-oriented pizza-box antennas and the emergent D2D play, Amazon Leo’s positioning is aggressively enterprise: deeply integrated with AWS, with “Direct to AWS” and “Private Network Interconnect” pitched as the primary connectivity models. Customers are told they can build private networks that never touch the public internet, with Leo acting as a secure, low-latency bridge between remote sites and AWS regions or enterprise data centres. It’s a very Amazon move: make the satellite terminal look like just another component in an end-to-end cloud architecture, rather than a standalone connectivity product.
For the satellite industry, Leo Ultra is less about the raw speed numbers and more about who it targets. The explicitly named early adopters include JetBlue, which plans to use Amazon Leo for “free inflight Wi-Fi,” along with energy companies and agricultural operators. This is the heartland of traditional GEO/HTS and VSAT: aviation connectivity, remote oil and gas, agritech, logistics. GEO operators have been spending the last decade pitching “cloud-ready” connectivity to this base; Amazon just skipped the dance and plugged the constellation directly into AWS.
It’s also hard to ignore the optics of Amazon cheerfully noting that it already has “more than 150 satellites in orbit” and is shipping Leo Pro and Leo Ultra terminals to customers as part of its preview. It’s a quiet way of saying: yes, Starlink is ahead, but Project Kuiper/Amazon Leo is out of the slideware phase and into the hardware-on-roof phase. For enterprises that already live and die inside AWS, sticking a Leo Ultra on a remote site and treating it as just another AWS attachment may be a lot more palatable than dealing with Starlink’s relatively siloed ecosystem.
From a competitive-landscape view, this week makes the emerging segmentation clearer. Starlink is leaning into mass-market and mobile-operator partnerships, especially via D2D and roaming-like integrations. Amazon Leo is carving out an enterprise-cloud niche with very high-spec terminals and tight AWS coupling. Eutelsat, SES, Viasat and friends are suddenly stuck in the uncomfortable role of the long-serving generals forced to share the war room with two very rich, very ambitious newcomers who are bringing their own logistics, supply chains, and app stores.
Eutelsat’s €1.5 billion recapitalisation: financing IRIS² and LEO while everyone complains
On the “GEO empire trying to stay relevant” front, the week’s most consequential development was financial rather than technical. Eutelsat pushed ahead with a €1.5 billion capital increase that is explicitly tied to funding its 2026–2029 investment plan, including accelerated deployment of its LEO assets and support for the EU’s IRIS² secure connectivity constellation.
The structure is a two-step manoeuvre. First, “Reserved Capital Increases” of about €828 million were completed on 21 November 2025, subscribed by the French State, Bharti Space, the UK government, CMA CGM, and France’s Fonds Stratégique de Participations Then, a rights issue for approximately €660 million was launched on 25 November, with subscription rights detached on 26 November and the subscription period opening on 28 November.
White & Case, advising the underwriting banks, very helpfully summarised the big picture in a 27 November press release: the combined €1.5 billion raise is part of a strategic roadmap “aimed at accelerating the deployment of its low Earth orbit (LEO) satellite activities and supporting the future IRIS² constellation,” while also reducing leverage. In other words, Europe’s only integrated GEO-LEO operator is passing the cap around so it can keep throwing serious money at LEO and at the EU’s multi-orbit sovereign connectivity project.
Outside the official documents, commentary has been… less charitable. Advanced Television noted on 28 November that Eutelsat shareholders are “upset” about the rights issue, highlighting that the week saw roughly €1.5 billion of fresh money raised with heavy dilution, partially to finance IRIS² participation. The prospectus itself acknowledges that the rights issue implies a steep discount versus the theoretical ex-rights share price, and spells out the dilution impact for shareholders who don’t participate.
Strategically, this week showcases both the strength and the fragility of the GEO-LEO hybrid bet. On paper, Eutelsat’s position is enviable: 34 GEO satellites, more than 600 LEO spacecraft via the OneWeb constellation, and a starring role in IRIS². In practice, the company is having to raise capital under pressure, with state shareholders stepping in as backstops, just as Starlink and Amazon Leo are scaling without public-market scrutiny.
There’s also a deeper political undercurrent: the French and UK states are effectively paying, via this capital increase, to ensure that Europe has at least one native GEO-LEO champion capable of playing in IRIS² and competing against U.S. constellations. That might be good industrial policy, but public markets have limited patience for multi-billion-euro “strategic autonomy” projects that show up as CapEx now and hypothetical incremental EBITDA in 2029.
If this were a historical epic, this is the scene where the treasury is emptied to fund both a new fleet and a land army, while the nobles grumble about taxes. The difference is that in this version, the invading force arrives in the form of cheap phased arrays and low-latency cloud on-ramps.
ESA’s CM25: Europe writes a very large cheque for secure, multi-orbit connectivity
While Eutelsat was soft-selling dilution to investors, European governments were having their own moment of fiscal courage at ESA’s Council Meeting at Ministerial Level (CM25) in Bremen on 26–27 November.
ESA announced that its Member States had committed a record €22.1 billion in subscriptions at the ministerial, a 32% increase in nominal terms over the 2022 ministerial. The official article, “ESA Member States commit to largest contributions at Ministerial,” makes clear that a significant chunk of this uplift goes into applications, including telecommunications, as part of a broader “European Resilience from Space” initiative bundling Earth observation, navigation, and secure connectivity.
The following day, ESA drilled into the satcom specifics with a dedicated release: “Europe chooses resilient and secure space-enabled connectivity with €2.1 billion investment,”. This piece confirms that over €2 billion in subscriptions have been committed to connectivity and secure communications programmes under the ARTES framework and the IRIS²-related secure connectivity programme.
ESA is careful to frame this as Europe’s answer to the “direct-to-device revolution,” promising continued investment in solutions that integrate ground and satellite networks and emphasising Moonlight and other infrastructure as part of a multi-domain “system of systems.” In plainer language: the ministers signed off on turning satellite connectivity into a central pillar of European security and industrial policy, with IRIS² as the flagship and ARTES as the sandbox.
For commercial operators, this is both a lifeline and a constraint. The lifeline is obvious: €2.1 billion of ESA-managed funding aimed squarely at the technologies they care about, from software-defined GEO platforms to multi-orbit terminals and NTN integration. The constraint is subtler: ESA and the EU are making it plain that they expect interoperable, standards-based systems with European non-dependence in key technologies. IRIS² is explicitly supposed to be “open and not based on proprietary solutions” and to support 3GPP NTN standards “to the maximum extent possible,” as earlier EU documentation has stressed.
If you are a LEO or MEO operator that was hoping to sell a closed, vertically integrated system into European governments, this week’s decisions suggest you may want to rethink that roadmap. Europe is essentially announcing that it wants a multi-vendor, multi-orbit secure communications fabric that it can plug its defence, civil protection and infrastructure operators into, with IRIS² as the backbone and ESA as the technology referee.
Cinematically, this is the council-of-realms sequence: lots of banners, a grand hall, and a long scroll of commitments. Just remember that translating €22.1 billion of political enthusiasm into actual contracts, launches and services will take years, and operators with real capacity in orbit now will have plenty of room to manoeuvre while Brussels and ESA work through the choreography.
OmanSat-1 and Airbus OneSat: reports of GEO’s death remain exaggerated
Just in case anyone thought GEO communications satellites were done, the Gulf quietly provided a counter-example this week.
On 24 November 2025, Airbus announced that its OneSat software-defined platform has been selected for Oman’s first national communications satellite, OmanSat-1. The press release, published in Airbus’s newsroom under the title “Airbus-built OneSat selected for the first Omani satellite,” notes that the satellite will be a fully reconfigurable, high-throughput Ka-band spacecraft serving Oman, its economic waters, and broader regions of the Middle East, East Africa and Asia.
The underlying government-to-industry relationship is spelled out in regional coverage: Oman’s Ministry of Transport, Communications and Information Technology signed a cooperation agreement with Airbus Defence and Space to design, manufacture and launch OmanSat-1, with national operator Space Communications Technologies (SCT) appointed as the project’s execution arm and future operator. Reports in local media and specialist outlets confirm that OmanSat-1 will be based on OneSat and operate in Ka-band, with a strong focus on digital sovereignty, secure national communications and extending broadband coverage to rural areas in line with Oman Vision 2040.
From a market perspective, OmanSat-1 is a textbook example of why GEO remains attractive for many states: one national HTS spacecraft, with modern digital payloads, can offer controllable capacity over a tailored footprint for decades. Add in a domestically anchored operator (SCT, wholly owned by Oman’s sovereign wealth fund) and you’ve got a neat bundle of sovereignty, industrial development, and long-term capacity that no foreign LEO constellation can really match politically.
The interesting nuance this week is that OmanSat-1 uses OneSat, the same flexible GEO platform Airbus has been pitching into European and global markets that are increasingly shaped by IRIS²-style thinking. Reporting notes that OmanSat-1 is the 10th satellite ordered under the OneSat line, confirming that demand for reconfigurable GEO HTS remains very real even as LEO constellations proliferate. From GEO’s perspective, this is the equivalent of switching from battleships to aircraft carriers: fewer hulls, more flexibility, and a heavy emphasis on software-defined capability.
The upshot is that the GEO vs LEO narrative remains deeply oversimplified. OmanSat-1 shows that for many countries, the right answer is “GEO first, maybe LEO later,” especially where regulatory control, national security, and predictable long-term capacity trump low-latency use cases.
Regulators and orbital house-keeping: ARCEP closes the L-band consultation, FCC signs off Intelsat retirements
In the regulatory trenches, this week didn’t deliver headline-grabbing spectrum wars, but there were some important undercurrents that directly touch D2D, MSS and GEO end-of-life operations.
On the European side, France’s ARCEP reached the deadline for submissions to its consultation on L-band mobile satellite services. While the consultation itself was launched in October, with a closing date of 21 November 2025, this week marks the end of the comment period and the start of the behind-closed-doors phase where ARCEP decides how to treat L-band in an era of IoT and D2D. The Bird & Bird summary of the consultation highlights ARCEP’s focus on emerging D2D services and IoT applications, alongside traditional satellite telephony, and emphasises that existing L-band authorisations for current operators expire between 2027 and 2030.
That sets up a very real collision course. On one side, incumbent GEO MSS operators have been using L-band for years to provide safety-of-life and narrowband services. On the other, new players (including D2D constellations and IoT-focused LEO fleets) are salivating over the prospect of harmonised spectrum that can be married to 3GPP NTN waveforms and integrated into smartphone modems. The close of ARCEP’s consultation this week means the clock just started ticking on a French (and by extension, EU) decision that will heavily influence how much room there is for D2D in L-band versus using other bands and more contorted sharing regimes.
Across the Atlantic, the U.S. FCC quietly reminded everyone that satellites eventually have to die somewhere other than the geostationary arc. A Public Notice document detailing recent satellite earth-station and STA actions shows a flurry of approvals for Intelsat License LLC between 20 and 24 November, including multiple 30-day special temporary authorities to operate TT&C facilities in support of deorbit and controlled drift maneuvers for Intelsat 11, as well as drift operations for the UK-licensed HYLAS-2 satellite. The same document authorises Kongsberg Satellite Services to operate a Maui ground station to support the VAST Haven-Demo mission, a reminder that on-orbit servicing is slowly graduating from conference panel topic to real RF-coordination problem.
Taken together, these regulatory moves show the other side of the D2D and LEO expansion story. While everyone is busy cheering new services and capacity, regulators are having to puzzle out how to rearrange spectrum allocations for MSS in an NTN world, and how to avoid turning GEO graveyard orbits into a permanent layer of unmanageable debris. The Intelsat STAs read like a checklist of late-career GEO fleet management: drift this one here, raise that one there, keep TT&C coverage while you nudge an old bird 330 km above the arc.
If you squint, this looks a lot like the end of a long campaign in a strategy game: you’ve filled most of the map, your early-game units are obsolete but still hanging around, and you now have to micro-manage retirements and upgrades while a new kind of warfare (LEO mega-constellations and D2D) is starting on another layer of the board.
LEO launch treadmill: Starlink’s 150th Falcon 9 of the year and China’s Shiyan-21
None of the above matters if there aren’t satellites in orbit. Unsurprisingly, this week also featured the usual drumbeat of LEO and GEO launches, with Starlink once again taking the prize for sheer volume.
In the early hours of 22 November 2025 (U.S. Eastern time), a Falcon 9 launched 29 Starlink satellites from Cape Canaveral, marking SpaceX’s 150th Falcon 9 mission of the year. Coverage in Space.com confirms the launch time (2:53 a.m. EST) and notes that SpaceX explicitly called out the 150-launch milestone in its post-launch messaging. The mission is part of the ongoing deployment of Starlink’s various orbital shells, including the Direct to Cell layer that underpins the Kyivstar launch discussed earlier.
The operational takeaway for the industry is the same depressing one we’ve had for a while: anybody trying to compete with Starlink on capacity, coverage, or refresh cadence is being forced to do so against a player that treats “150 orbital launches in a year” as a normal KPI. That launch ability is what makes announcements like the VEON and Proximus partnerships credible for Starlink; operators and regulators can reasonably assume that the spacecraft will actually be there.
China, for its part, continued its own low-profile but steady satcom build-out. On 21 November 2025, Chinese state media reported the successful launch of the Shiyan-21 communications technology experimental satellite from Xichang, aboard a Long March vehicle. Official coverage describes it as a test satellite for new communications technologies in orbit, with the usual lack of hard technical detail. The Shiyan series has historically been associated with on-orbit technology demonstrations, including potentially dual-use capabilities.
From the perspective of GEO/LEO commercial operators, these launches matter less for their immediate capacity and more for what they say about the global industrial baseline. While Europe debates IRIS² timelines and carefully structures rights issues, both the U.S. and China are quietly normalising a cadence of launches and experimental satellites that will define the rules of the road in LEO and MEO by sheer presence.
In pop-culture terms, think of this as the montage sequence where one faction keeps training in the background while the others argue over governance models. When the third act starts, the side that has been doing 150 launches a year tends to have more options.
The quieter but telling ripples: Orange’s D2D, Australia’s UOMO debate, and the shape of things to come
Strictly speaking, some of the initial D2D announcements from Orange landed just before our 21–28 November window, but their aftershocks were very much alive this week and help frame where the market is heading.
Orange’s 18 November press release made it the first European operator to launch a direct-to-device satellite SMS service (“Message Satellite”) in partnership with Skylo, initially for Google Pixel 9 and 10 devices on its 5G and 5G+ plans in mainland France. This week, tech and telecom media were still busy explaining the offer, dissecting pricing, and positioning it against rival plays like Starlink’s Direct to Cell and the emerging Proximus/BICS fabric.
In parallel, on the other side of the planet, the Australian government’s plan for a Universal Outdoor Mobile Obligation (UOMO) triggered a fresh round of operator pushback, including concerns about mandating satellite-based coverage via LEO networks such as Starlink before the ecosystem and spectrum planning are mature. While this is national policy rather than an orbital development, it illustrates the same pattern we see in Europe: governments are increasingly tempted to legislate satellite-assisted universality into being, while operators warn about costs, handset compatibility, and the small matter of relying on satellite providers that aren’t bound by the same obligations.
These ripples matter because they show how quickly D2D and hybrid satellite-terrestrial models are moving from “innovation demo” to “regulatory hot potato.” Between Orange’s consumer satellite messaging, Kyivstar’s wartime Direct to Cell launch, Proximus Global’s IPX partnership, and Australia’s UOMO debate, we’re seeing the outlines of a new normal where satellite is no longer just “backhaul for VSAT” or “maritime/aviation niche.” It’s an expected component of national mobile coverage plans, even for consumer handsets. MNOs that don’t have a D2D/NTN roadmap are at risk of being left on the defensive with regulators and customers.
The biggest frictions are about who carries obligations, how spectrum is allocated, and whether satellite operators can be compelled to behave like telcos when they’re providing quasi-critical coverage.
So where does that leave GEO, LEO, MEO, D2D and the rest after this week?
Looking back at the week of 21–28 November 2025, a few themes stand out once you strip away the press-release gloss.
If you zoom out, the industry this week looks like a slightly chaotic but undeniably forward-moving ensemble cast. GEO isn’t being written out of the script; it’s getting a character arc about reinvention and sovereignty. LEO isn’t just a scrappy side character anymore; it’s running multiple plot threads at once: consumer broadband, D2D, enterprise cloud, and experimental national constellations. MEO is lurking in the background, quietly providing O3b-style capacity and preparing to ride whatever standards ESA and 3GPP eventually bless for multi-orbit services. D2D has gone from “cool teaser trailer” to “live feature, in production, in a warzone.” And VSAT is evolving from being the star of the show to being one of several ways to plug into an increasingly hybrid network.
For an industry that has historically moved at the speed of orbital slots and ITU conferences, this week shows a surprising ability to behave like a proper tech sector: real services deployed, regulatory frameworks nudged forward, financing secured, and actual users benefiting from something other than just more TV channels. Yes, the capital structures are creaking, the regulatory timelines are glacial, and the risk of creating an orbital junkyard is non-trivial. But compared to a few years ago, when half the NGSO business plans looked like wishful thinking wrapped in constellations of buzzwords, this week felt refreshingly grounded.
We’re still a few seasons away from seeing whether GEO, LEO, MEO, D2D and IRIS² can share the same universe without turning it into an unwatchable crossover.
For now, at least, everyone’s still alive,
the budget just got renewed,
and the writers have clearly decided to raise the stakes.
