An approach to splitting 5G base stations between Earth and orbit could revolutionize mobile communications—making seamless connectivity possible even in the most remote regions.

In a major step toward global mobile connectivity, Fraunhofer IIS has introduced a new method to integrate satellites into 5G networks, even if those satellites can’t fully function as base stations. This innovation, part of the TRANTOR project, could pave the way for seamless communication in even the most remote areas on Earth.
Traditionally, mobile networks rely on terrestrial base stations. But with the evolution of 5G—and eventually 6G—the aim is to merge terrestrial and space-based communications into what are called non-terrestrial networks. These networks would use satellites to extend coverage far beyond the reach of cell towers. However, existing mobile standards weren’t built to handle the vast distances and latency involved in satellite communication.
Adding to the challenge, satellites—though increasingly equipped with intelligent processors—are constrained by the need to be cost-effective, robust, and energy-efficient. This limits their computing capabilities.
Fraunhofer IIS has addressed this challenge by developing a technique that “splits” a 5G base station. Instead of placing all processing on the satellite, only part of the signal handling is offloaded to space, while the rest stays on the ground. This significantly reduces the satellite’s processing and power requirements while still allowing it to play an active role in the network.
“This splitting method opens the door to flexible and scalable network architectures,” says Rainer Wansch, Head of the RF and SatCom Systems Department at Fraunhofer IIS.
The approach also supports future 6G networks, which will include not just satellites, but airborne platforms like drones and aircraft. During lab tests, Fraunhofer IIS simulated real-world space conditions using a channel emulator and established the satellite link via the DVB-S2X standard. A portion of the base station was run on an off-the-shelf FPGA, suggesting real-world applicability for future satellite deployments.