Beyond Connectivity

A: Sovereignty in space was a major theme at the Satellite 2026 show in Washington, D.C., earlier this year. The concept has rapidly shifted from asset ownership to controlling the entire operational stack: hardware, data, networks and operational autonomy. In this new paradigm, sovereignty is not necessarily tied to geography either. For example, Ukrainian officials reportedly concluded that sovereign data was more secure outside of Ukraine.

The EU is also evolving its definition of sovereignty quickly. The EU GOVSATCOM initiative, which went live in 2026, is a strong example of how government is driving commercial operators to pool together into a sovereign network. Likewise, global operators such as AST SpaceMobile are now willing to engineer sovereignty at the operational layer by opening an operations center in Germany with a command switch, effectively handing European partners control over encryption keys and beam management, not just access to them. Sovereignty has become the price of admission to European regulatory and political acceptance.

A: Mega-constellations and D2D services create a paradox for small and emerging space nations. On one hand, they can rapidly expand coverage and bring connectivity to underserved users and unconnected mobile devices outside cellular networks, both critical for an evolving digital economy. On the other hand, these networks are global, concentrate control over their users’ data and create dependency on another nation’s infrastructure. The barrier to access is low, and these capabilities exist in handsets and in orbit, whether or not the nation has formally licensed them.

A: As I noted in my Insight, Key Takeaways from Satellite 2026: NTN, Defense, and Sovereignty, the irony of the sovereignty push is that it cuts both ways for the market. Leasing commercial capacity alone is no longer viewed as sufficient, but every country building a sovereign network is also untenable. Hybrid architectures help solve this by layering sovereign-controlled assets (dedicated government systems) with nationally or regionally pooled commercial capacity and multi-orbit commercial services procured under sovereign-aware contracts. As a result, a middle-tier of sovereignty is achieved, backed by architecture and contracts rather than full-stack ownership.

A: The EU has been a stand-out here. GOVSATCOM, IRIS², and the EU Space Act collectively act as a strong foundation for a sovereign regulatory architecture. Together, these initiatives cover compliance, encryption and cybersecurity mandates as a “values-led” sovereignty rather than commercial-led or state-led.

Other good examples include small-nation models from Australia, Singapore and Luxembourg. Rather than building full sovereign constellations, they build enabling environments, incubators, policies and niche supply chains. This demonstrates sovereignty through value chain participation rather than full-stack ownership.

A: As satellite networks evolve towards more software-defined, agile systems in space, the parallel evolution of the ground segment away from fixed gateways toward more flexible, programmable orchestration layers is becoming critical. The ground segment is no longer a purely physical termination point, but a software layer that enables operators to mix different applications, waveforms, antennas and orchestration layers across different vendors. AI and machine-learning remain central in this evolution, unlocking capabilities like intelligent scheduling, predictive maintenance, anomaly detection and dynamic resource allocation at scale.

We can already see this shift happening commercially. ST Engineering iDirect’s Intuition platform, available since late 2025, uses cloud-native, microservices-based architecture that can reduce hardware requirements by up to 70%. iDirect also introduced a consumption-based service model, Intuition Unbound, which signals a broader shift from CAPEX-intensive infrastructure to OPEX-driven, on-demand microservices. On the 5G NTN front, Kratos Defense has been advancing has been advancing its OpenSpace software-defined ground system since 2025 with SES (formerly Intelsat) and was selected by JSAT in 2026 to develop and validate a 5G NTN ground system for APAC deployments using existing VSAT systems.

A: As multi-orbit constellations proliferate, the complexity of managing multi-orbit, multi-waveform networks can’t be orchestrated in real time without software-defined control layers. Current bent-pipe (transparent) architectures place a heavier orchestration burden on the ground, while emerging regenerative payload architectures offload some of that complexity to the satellites themselves as parts of the gNodeB move onboard. Most near-term commercial deployments transparent, however, meaning cloud native ground systems are the most economically and operationally viable path forward for multi-orbit networks, as they can avoid the rigidity of dedicated orbit-specific hardware stacks.

The ground layer can also bridge legacy satellite waveforms and 5G core. LEO’s high velocity also introduces additional handover challenges, from Doppler-aware scheduling to pre-compensation, which is driving demand for predictive and AI-assisted handover approaches that cloud-native ground systems are best positioned to enable. The near-term bottleneck isn’t in the 3GPP standards, but in the ground segment’s ability to become a true software control layer that speaks to both satellite and 5G core across orbits.

A: The biggest constraints aren’t technical but are structural. Legacy proprietary infrastructure, organizational silos between satellite and terrestrial operation, unresolved business models, fragmented spectrum harmonization across bands and the draw of vertically integrated close systems all remain real constraints. Vendors need to prioritize investing into open, software-defined orchestration layers that offer strong backward compatibility with legacy systems. Operators should also prioritize OSS/BSS convergence and consumption-based ground infrastructure models that let them scale NTN services without breaking the balance sheet on CAPEX-heavy builds before the revenue model is proven.

A: Some of the biggest remaining hurdles are latency and cost. Reliably delivering sub-100 millisecond latencies can unlock more in mobile device and IoT use cases beyond basic connectivity, and NTN-integrated chipsets need to hit price points that make mass-market devices viable. Until connectivity plans for NTN become more affordable, the business case outside of niche verticals like maritime and IoT remains fragile.

A: Most operators are pursuing a phased multi-vendor approach. MNOs are becoming satellite service aggregators, using NTN first as a supplemental coverage layer in rural or maritime gaps before building toward seamless terrestrial/non-terrestrial handover in IoT mobile connections and finally into the cellular subscriptions itself. Release 17 is foundational because it standardizes the NTN-specific physical layer adaptations that make satellite-native 5G possible but Release 18 and 19 will be where true service continuity matures. We are seeing the operators who are integrating NTN into their network architecture today have been able to increase geographic footprint 5–10% and population-coverage gaps another 1–3%.

A: Clearly first-mover advantage is showing advantages with gaining subscriber and building partnerships with Telcos, as mobile carriers hold the subscriber distribution and aggregate the satellite services. Chipset vendors, particularly Qualcomm and MediaTek, may ultimately be the market makers, since NTN adoption at scale depends entirely on how broadly and affordably they embed NTN support into mainstream silicon.

A: Currently, China has around five major mega constellations planned. These include China Satellite Network Group’s Guowang (13,000 satellites by 2035), Spacesail’s Qianfan (15,000+ satellites by 2030), and Hongqing Technology’s Honghu-3 (10,000 satellites). In January 2026, the newly established China’s Institute of Radio Spectrum Utilization and Technological Innovation filed for two additional mega constellations, CTC-1 and CTC-2, with a combined total of almost 200,000 satellites, demonstrating China’s long-term ambitions in the space industry. While there are few official details on the purpose of the orbital slots, the filings could be part of a broader strategy to reserve space for future commercial, military and security purposes, rather than leaving those positions open to competitors like SpaceX. Despite the scale of its plans, China currently lags in execution with only a few hundred satellites launched to date, compared with Starlink’s several thousand satellites already in orbit and operational.

In terms of China’s D2D strategy, satellite operators, state-owned telecom operators and the government are coordinated and working closely together to provide D2D services and satellite connectivity to their customers. For instance, China Unicom and China Telecom are already licensed to offer D2D services utilizing the state-owned Tiantong GEO satellite system. China Mobile uses the BeiDou navigation satellite system and plans to integrate with emerging LEO, MEO and GEO constellations to further expand D2D capabilities. Overall, the satellite and telecom industry are closely aligned with national technological priorities and policy objectives.

On the other hand, Western D2D strategies are largely driven by the private-sector and partnerships between mobile network operators (MNOs) and specialized satellite operators. For example, key partnerships include T-Mobile/Starlink, Verizon/AST SpaceMobile/ Skylo, Vodafone/AST SpaceMobile. In addition, their efforts are closely tied to the standardization of 3GPP Release 17/18/19 to enable broader device compatibility, accelerating commercial adoption.

We can see that China’s approach signals a long-term objective to build a large-scale sovereign space infrastructure that supports national digital infrastructure goals (such as integrated land-sea-air-space connectivity and 6G) to complement existing terrestrial networks and advances technological self-reliance.

A: China’s close alignment between satellite operators and state-backed telecom operators is demonstrating a highly coordinated, top-down approach. In late January 2026, China released its 15th five-year roadmap for its space sector, which sets out a unified commercial roadmap across various technological areas such as developing its orbital digital infrastructure (edge computing and AI platforms), satellite mega constellations and their integration with adjacent technological segments amongst other goals. This policy framework enables tighter coordination across stakeholders and ensures that satellite communications development aligns with broader national connectivity and industrial goals.

With strong alignment between satellite operators and telcos, China is treating space assets and infrastructure as an extension of the 5G/6G network, rather than standalone systems which supports a consolidation of spectrum usage, reducing fragmentation and enabling more efficient planning for large-scale satellite internet infrastructure. This integrated approach is designed to support seamless terrestrial and NTN interoperability within China in the long-term. In addition, Spacesail and other Chinese space networks are operating in regions (such as Asia, Africa and Latin America) without Chinese telco infrastructure. They do this by bundling services for enterprises customers – which are then delivered to consumers and end users through a B2B2C business model. This approach supports their commercial rollout beyond China.

The close alignment between both the satellite and telco industry is advancing 5G NTN standardization in China. Some key examples include ZTE and China Telecom completed a maritime 5G NTN trial where data transmission was successful via a GEO satellite. China Mobile has been pioneering 5G NTN field trials to verify terminal-to-satellite connections that focuses on integrating satellite communication with the 5G core network and supporting 5G-A initiatives.

A: A state-coordinated D2D ecosystem and China’s centralized approach to spectrum filings and mega-constellation planning can intensify competition for spectrum and orbital resources at the global level, particularly at the ITU. Large, early filings by state-backed entities such as China Satellite Network Group’s Guowang constellation and newer filings like CTC-1 and CTC-2, imply that spectrum access can be treated as a sovereign strategic tool rather than a commercial asset. China will be able to use a coordinated state approach to secure priority rights at the ITU, which can potentially block Western access in key regions, especially private satellite operators. Furthermore, a vertically integrated ecosystem threatens to bifurcate global standards. By controlling the entire stack – from the satellite chipsets and handsets to the constellations – state-backed players can subsidize hardware to undercut Western rivals. This creates a lock-in effect, where countries are forced to adopt Chinese-specific hardware to access the satellite network, effectively bypassing Western supply chains out of the loop entirely.

In terms of handsets, the state-driven ecosystem forces domestic manufacturers (like Huawei) to design handsets that support native, seamless switching between terrestrial mobile networks and satellite networks. However, these could limit global interoperability, especially for global device makers seeking to serve both the Chinese and Western markets. For example, Huawei smartphones already support satellite messaging via the BeiDou and Tiantong systems, delivering seamless functionality domestically but may risk limiting global interoperability if implementations rely on China-specific frequency bands and network architectures.

A vertically integrated model encourages domestic sourcing of materials, satellite platforms, payloads, launch services, and network equipment, which increases self-reliance and reduces dependence on Western or global suppliers. For Western vendors, this could mean heightened competition from Chinese suppliers in emerging markets (Africa, Latin America, and Asia-Pacific). In these regions, China’s ability to offer an integrated satellite-to-device ecosystem could challenge Western operators that rely on multi-partner deployment models.

A: The demand is driven by the collision of three exponential trends: the downlink bottleneck of Earth observation (EO), the energy crisis of terrestrial AI and the rise of strict “sovereign data” laws.

From an EO data perspective, we are generating petabytes of data in orbit, but the download throughput is a challenge. It could be potentially more efficient to process at the edge in space. Second, terrestrial grids are hitting a breaking point. Starcloud and Crusoe Energy are pioneering a model where high-density AI training loads are offloaded to orbit to harvest “stranded” solar energy stuck in orbit—effectively bypassing Earth’s carbon limits.

Finally, there is a regulatory arbitrage play. Axiom Space and others are exploring “orbital data vaults” that turn regulated data into “unregulated insight” before it ever touches a border. This allows a space-based data center operator to be not restricted by regional and local data center compliance but also take advantage of global connectivity.

In the long term, this looks like an ecosystem play. It’s not just about storage or processing; it’s about an orbital economy. SpaceX reduces the launch cost to deploy these assets. Starcloud and Crusoe are already partnering to offload high-density AI workloads to orbit to bypass the constraints of Earth’s terrestrial energy grid. Meanwhile, Intuitive Machines is building the lunar data relay network (NSNS contract) that will eventually pipe data from the lunar surface back to these orbital nodes, creating a continuous “Earth-Moon digital highway.”

A: To leverage the potential of physics. Space has “free” resources with continuous solar energy and passive cooling. Terrestrial data centers use a significant amount of power, with upwards of 40% or more of total energy consumed on active cooling (fans and chillers). In orbit, we can radiate waste heat directly into the cold void of space. While this requires significant radiator surface area, it eliminates the massive electricity cost of terrestrial cooling, theoretically offering higher energy efficiency.

I could see a number of use cases for these facilities from edge AI for EO and LLM training runs before being sent down to Earth, to sovereign compute (data vaults) like what Axiom space is exploring, to telemetry and navigation support lunar logistics like with Astrolab’s FLEX rover in support with Intuitive Machines data relay.

A: The zero-water advantage is highly realistic because the baseline on Earth is unsustainable. A medium-sized terrestrial data center cycles through millions of gallons of water annually for cooling. An orbital data center consumes zero water. There is also the energy perspective, which Google’s Project Suncatcher research suggests that a constellation of solar-powered satellites using optical links could effectively create a “virtual” gigawatt-scale data center without adding a single gram of carbon to Earth’s atmosphere.

Even so, we would need to evaluate the logistics of getting a substantial amount of data center capacity into orbit, if the objective is to operate data centers in space as a substantial complement to data centers on Earth. Nonetheless, data center computational capacity will be needed for in-orbit operations, lunar operations, and potentially even deep space operations.

A: Again, physics. I did a post on LinkedIn about this and the biggest hurdle isn’t getting there; its keeping the “chips” alive. The thermal management issue in particular is a challenge since dissipating heat is notoriously difficult for high-density GPUs. Furthermore, radiation—cosmic rays—rapidly degrades silicon, creating a shorter hardware cycle than on Earth. This necessitates in-orbit servicing models since maintaining the racks in orbit is a challenge. We anticipate a need for autonomous logistics vehicles—mirroring the model Astrolab is deploying for lunar surface logistics—to robotically swap degraded server components and maintain the constellation.

Despite these engineering hurdles, ABI Research identifies this as a critical growth vector in the New Space economy. We are tracking significant market momentum in this sector and will be publishing our comprehensive analysis on the orbital compute market in H1 this year.