Eutelsat and Skynopy scale software-defined ground network

Eutelsat and Skynopy are expanding their AKAR ground-segment work into a global software-defined network, with more than 100 operational antennas planned at selected sites by the end of 2029.

Global AKAR will combine Eutelsat’s existing ground infrastructure with Skynopy’s orchestration software and virtualised modem technology, giving satellite operators a common software interface through which to access antennas distributed across multiple locations.

The partners are targeting S-, X-, and Ka-band services across Earth observation, in-orbit services, orbital data centres, telecommunications, and other low Earth orbit applications. The supplied announcement also identifies defence and other sensitive missions as intended users, positioning Global AKAR as dual-use infrastructure rather than a network built around a dedicated military contract.

That distinction should remain clear. No defence customer, military procurement, or sovereign programme has been announced for Global AKAR, and its principal proposition is broader satellite infrastructure. Its relevance to defence lies in the same engineering problem faced by commercial constellation operators: data collected in orbit has limited value if it cannot be transferred to terrestrial networks quickly, reliably, and at sufficient volume.

The ground segment determines how often a satellite can make contact, how much information can be transferred during each pass, and how quickly that data becomes available to the user. A high-resolution sensor or capable onboard processor can still be constrained by infrequent station access or insufficient downlink capacity.

Global AKAR is intended to increase the number of available contact opportunities by presenting a geographically distributed pool of antennas through one orchestration layer. Eutelsat contributes existing sites from its global ground network, while Skynopy supplies the software intended to make selected infrastructure available across multiple satellite operators and missions.

The partners are targeting latency below three minutes, compared with what the supplied release describes as a current state of the art around one hour. Planned throughput reaches up to 10Gbps, against the 1–2Gbps comparison used in the announcement. Those figures remain programme targets rather than demonstrated network-wide performance.

The technical base is at an earlier stage. A three-antenna proof of concept has already been used to demonstrate feasibility for Earth observation, with part of the work supported by the French Government. The current partnership aims to turn that limited configuration into a global service spanning more than 100 operational antennas by the end of 2029.

Skynopy’s existing network provides another foundation. Published company material identifies 17 operational sites serving several dozen satellites, using a hybrid model that combines shared infrastructure with software intended to simplify access across different stations.

The industrial challenge is to make physically different ground stations behave like one service. Individual sites may have different antenna sizes, frequency capabilities, tracking systems, radio-frequency chains, modems, control software, terrestrial network connections, and maintenance arrangements.

Skynopy’s orchestration layer has to hide enough of that complexity for operators to request downlink services through a common interface while still respecting the limits of each site. An S-band antenna cannot become a Ka-band gateway through software alone, and throughput remains constrained by the physical RF chain, spectrum allocation, and terrestrial backhaul available at each location.

Virtualised modem technology gives the software layer greater flexibility. Moving modem functions from fixed dedicated hardware into software can allow processing resources and waveforms to be assigned more dynamically between users, reducing the amount of single-purpose ground equipment tied to one satellite operator.

That can improve utilisation of existing infrastructure, although it also increases dependence on compute availability, cybersecurity, software configuration, and network management. A software-defined ground segment therefore exchanges some hardware rigidity for a larger integration and assurance burden in the digital layer.

For defence and institutional missions, resilience becomes particularly important. A geographically distributed network gives satellite operators more potential contact points and reduces dependence on one station, but geographical spread alone does not guarantee resilient service.

Power, terrestrial connectivity, physical security, spectrum availability, cyber protection, and the ability to reroute traffic all affect whether another ground station can take over when one site becomes unavailable. The orchestration platform itself also has to avoid becoming a central point of failure as more physical infrastructure is brought under one software layer.

The inclusion of X-band broadens the network’s relevance to government and institutional users, while the supplied release states that secure architecture and interfaces are intended to make the system suitable for defence and other sensitive applications. That creates a potential alternative to building a dedicated global chain of sovereign ground stations, provided the commercial infrastructure can satisfy the customer’s assurance and security requirements.

Skynopy already cites Airbus Defence and Space among users of its ground-station services, and the company has previously published work supporting the Pléiades Neo Earth-observation constellation. Those activities provide operational evidence for the underlying service model even though Global AKAR itself remains in the scale-up phase.

The network is also being designed around changes in the satellite market above it. Larger LEO constellations and higher-resolution sensors generate increasing volumes of data, while more processing in orbit can change what information needs to be sent to Earth and how urgently it must arrive.

Orbital data centres would push the requirement further. Global AKAR identifies in-orbit computing among its intended applications, effectively treating the ground network as the terrestrial connection for processing infrastructure located in space. Latency, throughput, and software-based resource allocation then become more similar to digital-network requirements than traditional telemetry alone.

That same architecture is relevant to defence ISR and communications systems because operational value increasingly depends on shortening the path between collection and use. Commercial ground infrastructure could offer additional contact opportunities or network diversity, but sensitive missions would need clear controls around encryption, data handling, sovereignty, access priority, and service availability.

The partners have not disclosed how those defence-specific assurance requirements would be implemented, so the current announcement should not be read as confirmation that Global AKAR already meets military accreditation standards. The proposition is that the architecture is being designed to support sensitive users; actual customer deployments will determine how that claim translates into operational security requirements.

The commercial model is based on extracting greater value from existing infrastructure rather than building an entirely separate global antenna estate. Eutelsat supplies geographic reach and operating experience, while Skynopy provides the software layer intended to expose selected antenna capacity to additional customers.

Reaching more than 100 operational antennas by 2029 will still require substantial integration. Selected sites have to be connected into the orchestration platform, validated across the required frequency bands, linked to sufficient terrestrial backhaul, and operated through consistent cybersecurity and service-management processes.

If that deployment reaches the stated scale, Global AKAR would move the ground segment closer to shared digital infrastructure: operators would consume distributed antenna capacity through software without owning every physical site underneath it. For defence users, the unresolved question is how far that commercial flexibility can be combined with sovereign security and assurance requirements.

The partnership establishes the infrastructure route but not the military customer. Its significance to IN Defence lies in the ground segment becoming a software-defined, high-throughput layer that could support sensitive space services at much greater scale — provided the promised performance, resilience, and security survive the transition from a three-antenna proof of concept to a global operational network.