en.Wedoany.com Reported - The University of Electronic Science and Technology of China (UESTC), in collaboration with the Tianfu Wireless Intelligent Research Institute, has proposed a low-complexity digital beamforming method for low-earth orbit (LEO) satellite beam scanning. This method significantly reduces processing time while maintaining acceptable beam performance. Addressing the challenge of tracking ground targets due to the high-speed movement of LEO satellites, it provides a solution for real-time beam scanning communications.
Phased array antennas are standard in LEO satellite communications. Their large-scale arrays control the feed phase and amplitude of each element to change the direction of the pattern's maximum, enabling beam scanning. However, the high-speed motion of satellites causes ground terminals to constantly perceive changes in satellite position, necessitating a beam scanning method that is both fast and capable of maintaining multifunctional pattern performance to ensure communication continuity.
The method comprises two stages: first, a convex optimization process determines the excitation related to the initial beam; then, a low-complexity beamforming method determines the excitation related to the scanning beam. This process is formulated as an orthogonal Procrustes analysis problem with a closed-form solution, requiring no iteration or matrix inversion, thus achieving low-complexity computation. The method can rapidly synthesize complex scanning beams in specified directions, operating with acceptable beam performance loss.
Compared to traditional phase-shift methods, this approach better maintains complex beam shapes, achieving a favorable trade-off between beamforming performance and complexity, making it suitable for multi-user satellite communication scenarios. Numerical simulations in various scenarios validate its effectiveness: on a 16-element uniform linear array, the computational delay is as low as 0.01 milliseconds; on a 14×14 element uniform planar array, the computational delay is 0.18 seconds, while achieving acceptable sidelobe level degradation. Compared to improved genetic algorithms and convex optimization-based methods, this approach reduces processing time by 70% to 99.71% while maintaining strict sidelobe level performance bounds. Compared to traditional phase-shift methods, processing time increases by up to 25%, but the sidelobe level can be reduced by at least 76%.
The related research results have been published in the journal Science China Information Sciences (English edition) (Sci China Inf Sci, 2026, 69(8): 189304), co-authored by Tian Yutong, Yang Haining, Yi Shijia, Li Na, and Cheng Yujian from UESTC, and Yang Haining from the Tianfu Wireless Intelligent Research Institute.
