Submodular Optimization for Discrete Movable Antenna Positioning
Published:
City University of Hong Kong · Advisor: Prof. Xianghao Yu · 2024 – 2025
Movable (fluid) antennas can reshape the wireless channel by repositioning radiating elements, but selecting positions from a discrete grid is a hard combinatorial problem. This project develops a principled optimization foundation for it.
Contributions
- Proved that the discrete movable-antenna (MA) positioning problem is monotone submodular subject to a 2-system constraint.
- Proposed the DCSPS algorithm with a theoretical 1/3-approximation guarantee relative to the optimum.
- Extended the framework to imperfect-CSI scenarios via a robust DCSPS scheme.
Related publications: JSAC 2025 (under review) · ICC 2026