NTT and Safran Want Satellite Laser Links to Behave More Like a Network

A satellite can send data through a laser link, but the ground network still has to catch it. A telescope must point at a moving spacecraft, maintain a narrow optical path through the atmosphere, and hand the data into the terrestrial network that carries it to users. That makes the ground station and its control software as important as the laser terminal in orbit.

On November 6, Japan’s NTT said it had signed a memorandum of understanding with Safran Data Systems to explore next-generation optical ground communications for multiple satellite orbits. The proposed work would study how NTT’s digital coherent and network-control technologies could be integrated with Safran’s optical ground stations and connected to NTT’s All-Photonics Network (APN). The companies describe the aim as a path from low Earth orbit (LEO) through geostationary orbit (GEO) into terrestrial data infrastructure.

The wording matters. This is an agreement to examine a possible collaboration, not an announcement that a multi-orbit laser network is already deployed or that commercial service has a launch date.

The ground station is part of the link

Radio-frequency satellite links are the established option for many missions. Optical communications use tightly directed beams of light, which can support high data rates and reduce radio-spectrum constraints. But the beam has to be acquired and tracked with precision. At the ground link, clouds and atmospheric conditions can interrupt optical visibility; a satellite’s orbit also determines when a particular station can see it.

Safran’s existing IRIS optical ground station is designed for LEO, GEO and deep-space applications. The company says the system supports bidirectional links and integrates its Cortex Lasercom modem, with rates up to 10 gigabits per second in specified configurations. Those are product capabilities described by Safran, not a performance promise for the proposed NTT collaboration.

A network of ground sites can help address the visibility problem by offering more opportunities to acquire a satellite, but that introduces another engineering layer: moving sessions between stations, managing routing and keeping service behavior predictable as the spacecraft crosses coverage areas. NICT’s overview of optical ground communications describes the need to track satellite position and acquire weak optical signals automatically. Commissioning work on Australia’s TeraNet network likewise treats multiple sites and operational handoffs as a system, not simply a telescope specification.

What NTT brings to the proposal

NTT’s stated contribution is not a new satellite. The company wants to apply digital coherent communications and network-control technology to the ground segment, then link optical stations to APN. In broad terms, coherent optical systems encode information in properties of the light wave and use digital processing to recover it; network control determines how links and capacity are organized end to end.

That combination could matter if it helps manage changing satellite-to-ground paths as a continuous service. A data stream from a spacecraft would need to travel through the optical terminal, station equipment, terrestrial transport and onward to a data center or customer. If each segment is engineered in isolation, the fastest optical hop may still wait on a poorly coordinated terrestrial path.

The practical test is therefore not just peak link speed. It is how a system handles acquisition, weather-related outages, station handoff, buffering and recovery while maintaining a usable service-level target. For operators, a predictable lower rate or alternate path can be more valuable than an impressive maximum that is available only under ideal conditions.

The work still has hard limits

The announcement does not specify a deployment location, ground-station count, orbiting customer, project budget, technical milestones or service date. It also does not claim that APN will make optical links immune to atmospheric interruption. Ground networks generally need operational redundancy or an alternate communications path for periods when the optical path is unavailable.

For now, the useful signal is strategic: NTT and Safran are exploring a meeting point between satellite optical terminals and terrestrial network control. Safran has a product in this category; NTT has network technologies it wants to bring to the integration. Whether the collaboration progresses from memorandum to a tested system will depend on engineering details the announcement does not yet provide.

Readers can follow the next announcement for concrete evidence: a named test site, an on-orbit or ground demonstration, defined handoff behavior, measured availability, and a clear distinction between optical line rate and end-user throughput. Until then, this is an infrastructure direction under study, not a new service to order.

For a closer look at the terrestrial photonics behind future networks, see our coverage of an isolator-free photonic transmitter and optical I/O for AI clusters.