A mathematical model of satellite-to-ground wireless power transfer showed different scaling patterns for two coordination schemes. With coherent coordination, meaning the satellite signals are aligned, modeled received power followed a quadratic relationship with the number of satellites. With non-coherent coordination, it followed a linear relationship, and the comparison showed only a modest benefit from additional antennas.
The work is a preprint identified as arXiv:2608.25589v1 and dated 26 August 2026. It reviews microwave-based long-distance wireless power transfer and introduces coordinated space-based power transfer for networks of ground devices. Its evaluation covers average received power under coherent and non-coherent coordination, along with spillover at undesired locations.
A controlled satellite setup
The main distributed-power example placed satellites at the vertices of a regular polygon with a 10 km circumradius, the distance from the centre to each vertex. The modeled satellites were at an altitude of 500 km, operated at 12 GHz and had a maximum power budget of 1,000 W each. Five single-antenna ground devices were distributed within a circular area with a 2 km radius.
The satellite-to-device links were modeled as narrowband and line of sight. The model jointly optimized the satellites’ array precoders, the settings used to shape each transmitted signal, while keeping each satellite within its individual power budget. Its objective was to maximize the power delivered to the worst-served device.
The modeled power stayed in the milliwatt range
Across the modeled configurations, received radio-frequency power, the power carried by the incoming radio signal, ranged from sub-milliwatt levels with a few non-coherently coordinated satellites to tens of milliwatts under coherent coordination with larger arrays and more antennas. The article compared these levels with representative requirements for outdoor inventory, remote monitoring and an AI application.
For outdoor inventory, the representative profile listed an average direct-current power need of 0.60 mW and required radio-frequency input of 1.20 mW. Remote monitoring was listed at 1.25 mW DC and 2.50 mW RF. The AI application was listed at 2.09 mW DC and 4.18 mW RF. The emergency profile had no reported total DC or RF power.
The article says modeled configurations delivering 1.20 to 2.50 mW could theoretically support outdoor inventory. At least 2.50 mW could also support remote monitoring, while more than 4.18 mW could meet the average requirements of all three listed applications. These are theoretical feasibility comparisons rather than demonstrated device-charging performance.
Coordination and precision
The model also examined spillover, meaning power reaching locations outside the intended devices. The modeled results showed stronger energy concentration at target devices under coherent coordination. Configurations with more antennas and satellites had less spillover under both strategies. The supplied analysis gives no numerical spillover estimate.
The modeled coherent result was lower at higher residual phase-error standard deviations. Residual phase error is the mismatch left after signals are synchronized. The text does not report the size of the reduction, and the analysis says a practical system could operate in a partially coherent or non-coherent regime.
The conclusion treats changing propagation delays, Doppler shifts and the geometry of the satellites serving a device as beam-management challenges. It links inaccurate compensation to synchronization and beam-steering errors, lower gain or unintended nulls, where the signal is suppressed.
Other pieces of the proposed system
A separate model examined laser wireless power transfer between satellites. In that model, pointing loss, the loss associated with a misdirected narrow beam, first decreased and then increased as the transmitting aperture grew. It also increased with link distance, and the results indicated an aperture that minimized pointing loss.
The article also presents two possible roles for metasurfaces, engineered surfaces that shape electromagnetic waves. They could be integrated into satellite transmitters for lower-complexity beamforming or deployed near receivers to address non-line-of-sight conditions and enhance energy-harvesting efficiency. These are proposed uses within the framework.
A feasibility test, not a field result
The results come from a simplified model. The main coordinated-WPT results use narrowband line-of-sight links and a single network snapshot, and the analysis says the channel description did not include atmospheric attenuation, scattering, turbulence or shadowing.
The reported thresholds are theoretical comparisons with representative power requirements. They describe what the modeled setup might deliver under its assumptions, not what a real device is guaranteed to receive or what an end-device charging test has measured.
The work was partially supported by the Research Council of Finland through Grants 348515 (UPRISING) and 369116 (6G Flagship).
Paper data and sources
Original title: Power from Space: Coordinated Satellite Charging for Off-Grid Wireless Systems
Authors: Osmel M. Rosabal, Amirhossein Azarbahram, Mateen Ashraf et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-26
DOI: Not available
Original paper · Full text