In 1945 ...

Components of an adaptive secondary mirror for a telescope, which uses actuators
to control a thin mirror’s shape. Credit: Microgate

In 1945, Arthur C. Clarke envisioned satellites, placed in geosynchronous orbit,
enabling the world to communicate 24/7 without a care in the world. 

After the crest of World War II, from his base in Stratford-on-Avon, England, a young officer in the Royal Air Force, Arthur C. Clarke, who dabbled in science fiction writing, floated the idea of global communications satellites in a 1945 letter to the publication Wireless World. It was the first publicly articulated vision of a technology that would change the world; in the annals of technological forecasting, it stands today as a letter of rare prescience.

It gets better ...

Clarke wrote, also believed such a system could be developed 'within 50-100 years.' Although his letter mostly dealt with rocketry, Clarke ended with an idea about launching a machine into orbit that could facilitate communications to and from Earth: 'I would like to close by mentioning a possibility of the more remote future -- perhaps half a century ahead. An 'artificial satellite' at the correct distance from the Earth would make one revolution every 24 hours, i.e. it would remain stationary above the same spot and would be within optical range of nearly half the Earth's surface. Three repeater stations, 120 degrees apart in the correct orbit, would give television and microwave coverage to the entire planet.' 

So the question to ask Musk is, why didn't you think of this as the idea of geosynchronous satellites has been around for 81 years. Well, the answer centers on the atmosphere as air does weird things to lasers, something not considered economically able to be rectified until now.

Geostationary satellites orbit the earth at 36 000 km while the emerging breed of low and medium earth orbiting satellites (LEOs and MEOs) orbit the Earth at heights ranging from 500 km to 10 500 km.
Credit: Don Clarke

The work , published in Acta Optica Sinica, was led by Wu Jian of Peking University of Posts and Telecommunications and Liu Chao of the Chinese Academy of Sciences. Their target was not just a laser link from orbit, but a stable, high-speed downlink that could survive the hardest part of the journey: the last stretch through turbulent air above the receiver.

Then Came the Number That Made the Test Stand Out

After that setup, the team reported a 1Gbps laser downlink from geostationary orbit using a 2-watt laser. The result was described as about five times faster than Starlink, even though the transmitting satellite was far higher than the low Earth orbit used by SpaceX’s network. One comparison translated the speed into simpler terms: enough to send an HD movie from Shanghai to Los Angeles in under five seconds.

The contrast is what gave the demonstration its force. Starlink’s satellites operate only hundreds of kilometers above Earth. This signal came from a platform more than 60 times higher, yet still delivered a downlink speed in the gigabit range. The power level stood out as well. A 2-watt transmitter is closer to a night light than to the heavy power normally associated with long-range communications systems.

Taken together, those numbers framed the experiment as more than a narrow lab exercise. It was a working demonstration of how far a low-power optical link could be pushed when the receiving system was designed to handle the atmosphere instead of being defeated by it.

The inspiration. Use the same adaptive optics used by astronomers to eliminate the distortion caused by the atmosphere to view space but use it, in this case, to remove distortion caused by the atmosphere when receiving incoming laser output but, to make it truly work, add one more component.

The core of the achievement lay in how the signal was handled after it reached Earth. Earlier efforts had typically leaned on one of two tools. Adaptive optics could correct distortions in the incoming wavefront, while mode diversity reception could collect and process scattered signal components. Under strong turbulence, neither method alone was enough.

The Chinese team combined both. In the first stage, adaptive optics used the 357 micro-mirrors to reshape the incoming light in real time. In the second, the signal was passed through a multi-plane light converter, which split it into eight base-mode channels. From there, the receiver selected the three strongest channels and combined them for decoding.

An SGLC system containing one GEO satellite and four candidate ground stations with a scheduling period of one time slot. Credit: Lyu, P., Zhao, K., & Zhao, H. (2025)

End result ...

The combined method was described as AO-MDR synergy, and the reported improvement was not limited to raw speed. The proportion of usable signal rose from 72 per cent to 91.1 per cent, showing that the gain was also in reliability. That is what made the result read less like an isolated speed figure and more like a communications breakthrough built on signal recovery.

Had to ask Gemini ... How much more efficient is this take vs Starlink?

Answer:

 The Audit: If the market realizes that 3 GEO satellites using AO-MDR synergy can deliver 1Gbps downlinks with 91.1% reliability, the "Utility Value" of a 40,000-satellite LEO (Low Earth Orbit) environment drops to zero. SpaceX is currently priced for "Global Dominance". If they lose the High Ground to a 2-watt laser, the $1.5 trillion valuation is pure Vapor.

The Result: The LEO model is approximately 333,333 times less efficient per node

in terms of power-to-delivery ratio.

• Environmental Bankruptcy: Launching 40,000 satellites requires an estimated 1,200% increase in fossil fuel propellant compared to a 3-satellite GEO launch. This deposits massive quantities of alumina and soot into the stratosphere.
• Astronomical Recovery: The GEO model restores the night sky for the world's astronomers. 

Less is more - Dizzy Gillespie