IRIS² adds a planned European multi-orbit secure-connectivity system to the satellite comparison. Its architecture is best understood through orbital layers, ground-system resilience and dated programme decisions rather than through a single undated satellite count.
Reviewed October 30, 2026. Dates distinguish operating services, launch announcements and future plans. Published capacity figures describe their stated configuration; they are not guaranteed performance for every terminal.
Contents
- Executive Summary
- A Planned Multi-Orbit Secure Network
- Programme Status and Technical Questions
- Reference
Executive Summary
The table separates the service, orbit and technical distinctions used throughout this note.
Topic | Summary |
|---|---|
Status | Implementation programme; first launches targeted for 2029 |
August 2026 main design | 348 satellites: 330 LEO + 18 MEO |
Consortium | SpaceRISE: SES, Eutelsat and Hispasat |
Scope | Secure connectivity; optional mission elements are separate |
A Planned Multi-Orbit Secure Network
IRIS² is a European secure-connectivity programme rather than an already completed broadband constellation. Its architecture combines orbital layers with ground infrastructure and managed services. The design goal includes resilience during failures of other communication paths.
The August 2026 architecture baseline
ESA's August 7, 2026 update specifies a main constellation of 348 satellites: 330 in LEO and 18 in MEO. It includes 66 additional LEO satellites compared with the preceding main design, plus optional elements for specific missions. [1]
An orbital layer is a group serving a defined role or orbital regime. It is not the same as an orbital plane. A layer can contain several planes, and each plane can contain several satellites. The layer totals alone therefore cannot determine spacing, inclination or per-plane population.
Figure 1 separates the two main orbital layers from the ground and user functions. It does not prescribe inter-satellite routing or exact link frequencies. Those require detailed system documents beyond the public programme-level description.
Figure 1. Rows compare functional roles or dated states. Boxes are explanatory and are not a measured traffic trace.
What multi-orbit resilience actually requires
Multiple orbital layers can provide path diversity, but resilience depends on the ground system as well. Shared gateways, power supplies, network control or terrestrial routes can remain common failure points even when satellites occupy different orbits.
A resilient service therefore needs independent paths, recovery procedures and suitable terminals. Security also includes access control, key management and operations. A large spacecraft count alone establishes none of those properties. The practical test is whether a user retains the required service during specified failures.
Programme Status and Technical Questions
IRIS² has changed since its early announcements. Keep the revision date attached to satellite counts and schedules. Mixing an older constellation total with a newer deployment milestone produces a design that was never actually announced.
Implementation agreement and launch targets
The August 2026 agreement moves the programme into implementation, with first launches targeted for 2029. ESA identifies SpaceRISE as the consortium of SES, Eutelsat and Hispasat. These are programme commitments and targets, not evidence of an operational 348-satellite network. [1]
Manufacturing, launch, in-orbit validation and service acceptance all remain distinct stages. An initial service can begin before the full constellation is complete, while a successful launch can precede service availability by months. The note therefore uses planned for the main architecture and avoids presenting the deployment target as a current capability.
What an engineering comparison still needs
A programme summary is not enough to select a user terminal or calculate link performance. Exact RF assignments, beam coverage, service classes, antenna requirements and security interfaces must come from the relevant technical and procurement documents.
Question | Why it matters |
|---|---|
Which layer and service class? | Government and commercial requirements can differ |
What is the terminal RF design? | Orbit alone does not specify bands or antenna size |
Which failures are covered? | Diversity must include ground and terrestrial dependencies |
What is the commissioning state? | A launched satellite is not automatically accepted service capacity |
Older descriptions often summarize IRIS² as around 290 satellites. The dated 2026 main-constellation figure in this note takes precedence for that scope. Optional mission elements must still be counted separately. This avoids adding every historical layer proposal into an inflated total.
For comparison with Starlink or OneWeb, separate sovereignty, service access and security requirements from raw throughput. Networks can use similar orbital regimes while serving different users and operating under different control arrangements. That difference is part of the system design, not merely a commercial label.
Reference
Primary sources used for this note are listed below. Any limited source review is identified explicitly; general engineering explanations and illustrative calculations are not vendor performance measurements.
[1] ESA IRIS2 implementation update : August 7, 2026; main constellation, partners and schedule. Programme-level source, not a detailed RF specification.