World's First Successful 2 Tbit/s Free-Space Optical Communication Using Small Optical Terminals Mountable on Satellites and HAPS

The small optical communication terminals developed by NICT consist of two types: the high-performance FX (Full Transceiver) and the simplified ST (Simple Transponder) (see Figure 2). In 2024, NICT completed the development of the first practical prototypes used in this experiment. These terminals are designed to flexibly adapt to various platforms and scenarios through their component configuration and internal adaptive operation, allowing for the selection of an appropriate basic configuration depending on the application. The ST type is intended for small platforms or simple operations, while the FX type is selected when high capacity and high performance are required. Furthermore, depending on communication requirements, a 10G modem type (10 Gbit/s or less) or a 100G modem type (100 Gbit/s or more, capable of handling the Tbit/s range by using multiple channels via wavelength division multiplexing) can be selected to meet a wide variety of requirements. Once the basic configuration is determined, the terminal automatically adapts its performance according to fluctuating link conditions via internal adjustment functions.

To fit the terminals within the strict size, weight, and power consumption (SWaP) constraints of CubeSats, and to handle the dynamic environments assumed for mobile operation, the following three approaches were implemented:

· Development of Newly Designed Components: Examples include a 9 cm telescope (FX terminal) that achieves excellent optical performance in a compact form, and a novel configuration integrating a 3 cm telescope with a gimbal (ST terminal). Additionally, to handle dynamic environments assumed for mobile operation, NICT uniquely developed Beam Divergence Control (BDC) technology capable of dynamically adjusting the laser beam divergence according to link conditions. Currently, this technology is implemented only on NICT’s terminals. One of its features is that it facilitates laser pointing, addressing the challenge of narrowed beam widths typical in short-range communication, thereby ensuring a stable connection.

· Redesign and Modification of Commercial Components: Commercial MEMS mirrors were improved to withstand high optical power even in a vacuum while leveraging their characteristics that allow for miniaturization. Conventionally, it was necessary to use larger mirrors or lower the optical power, but these challenges have been overcome by NICT’s novel design. This is implemented as part of the high-precision alignment function using coarse tracking (gimbal) and fine tracking (MEMS mirror) to handle dynamic environments assumed for mobile operation. In addition, the optical amplifiers have been miniaturized and adapted for use in space environments based on commercial products.

· Active Utilization of Commercial Components: Commercial components are also actively utilized after verifying that they operate reliably in the space environment. A representative example is the high-speed transceivers adopted to miniaturize the modem.

The 2 Tbit/s modem (see Figure 3) was designed with a compact form factor (a single 20×20 cm board) so that it can be used in field experiments in combination with ST and FX terminals. This modem multiplexes 5 channels of 400 Gbit/s to achieve a maximum transmission rate of 2 Tbit/s on the same transmission path. Each channel is capable of operating at a maximum of 400 Gbit/s, but if the link conditions deteriorate, the data rate can be reduced to 100 Gbit/s. Optical signals for each channel are multiplexed into a single optical fiber by the modem and transmitted after high-power amplification. On the receiving side, the received weak optical signals are pre-amplified, and then the five channels are separated to remove noise and improve the signal-to-noise ratio.

To achieve miniaturization of the 2 Tbit/s modem, NICT applied methods demonstrated in the previous 10 Gbit/s modem. The conventional method of combining individual components increases the size of the modem, while integrating the entire modem using Photonic Integrated Circuits (PIC) requires significant investment and time. Therefore, NICT adopted a hybrid approach utilizing small, high-performance components used in terrestrial optical communication, effectively incorporating high-speed optical transceivers developed for data centers into this small free-space optical communication modem, thereby achieving both miniaturization and high performance.

This experiment was conducted by installing the FX terminal at NICT and the ST terminal at an experimental site 7.4 km away, with different types of terminals at each end. To significantly increase the transmission speed compared with the demonstration experiment using a 10 Gbit/s modem conducted in 2024, the 2 Tbit/s modem was selected for this experiment. The primary function of the terminals in this experiment is to maintain high-precision alignment between the terminals. This is realized by a coarse acquisition and tracking system (2-axis gimbal) and a fine tracking system (MEMS mirror). Additionally, the terminals perform high-power amplification before transmission (compensation for channel loss) and low-noise amplification after reception (raising the received signal power level to an allowable range), filtering background noise by removing out-of-band components.

The external dimensions of the ST terminal and FX terminal used in the experiment are 35 × 28 × 28 cm and 40 × 28 × 28 cm, respectively, with masses of approximately 13 kg and 28 kg; they are designed to be compact enough to be mounted on HAPS or satellites. Generally, HAPS payloads are constrained to dimensions of approximately 30-50 cm, mass of 5-25 kg, and power consumption of around 50-200 W; the ST and FX terminals generally meet these constraints while being equipped with a 2-axis gimbal. Furthermore, additional miniaturization is underway for the development of terminals for satellites. The optical head itself has already been miniaturized (see Figure 2), and the heaviest subsystems common to both terminals—control electronics, optical amplifiers, and power management units (shown as the square box under the optical head in Figure 1)—have already been miniaturized for the satellite-mounted version (not yet integrated into the current ST/FX prototypes). As a result, the mass of the entire satellite terminal is approximately 4 kg (excluding the gimbal), with the goal of fitting it into half the space of a 6U CubeSat scheduled for launch in 2026. Through this initiative, the new ST/FX terminals are expected to achieve a reduction of over 50% in both occupied volume and mass.

Horizontal free-space optical communication experiments at ground level are conducted under realistic atmospheric conditions prior to demonstration in space. In recent years, successful communication experiments exceeding 100 Gbit/s have been reported. In Asia, there were no reports of Tbit/s-class free-space optical communication experiments prior to NICT's experiment, although a Chinese research team achieved 120 Gbit/s over a distance of 1 km in Changchun in 2018, 100 Gbit/s over a distance of 2.1 km in Beijing in 2021, and single-channel communication of 112 Gbit/s over a distance of 104.8 km at Qinghai Lake in 2025.

In Europe, the German Aerospace Center (DLR) conducted an experiment of 13.16 Tbit/s total over 54 channels at 10.45 km in 2019. In 2023, ETH Zurich succeeded in 1 Tbit/s communication on a single channel over 53 km crossing the Swiss Alps. In addition, 4 Tbit/s in Aveiro, Portugal (2023), and 5.7 Tbit/s in Eindhoven, the Netherlands (2025), have been reported.

In all these Tbit/s-class horizontal links, the modem systems were configured by combining discrete components such as laser light sources, modulators, and receivers in a laboratory-style setup centered on rack/benchtop measuring instruments such as large arbitrary waveform generators and high-performance sampling oscilloscopes. Coherent signal processing and FEC decoding were often performed by analyzing acquired waveforms offline with external software, using numerous optical fiber patches and multiple racks of high-power equipment; thus, mounting on mobile platforms with size, weight, and power (SWaP) constraints is not realistic.

Furthermore, the free-space optical transmission sections were also configurations where commercial telescopes and optical components were assembled in a laboratory-style setup at fixed rooftop stations, and did not support environments where the platform moves dynamically. Therefore, these remained merely proofs of concept and were not systems suitable for the purpose of supporting Non-Terrestrial Networks (NTN) by mounting on mobile platforms like HAPS and satellites.

The 2 Tbit/s communication over 7.4 km achieved by NICT in this experiment is the first realization of long-distance and ultra-high-speed communication using small and lightweight terminals and modems, representing a major step forward toward practical application.

A communication method that transmits laser light through space. It is applicable to high-speed communication on the ground, between the ground and space, and in outer space.