AI & ML
When to Move: Online Repositioning of Movable Antenna Arrays Under a Long-Term Actuation Energy Budget
arXiv eess.SP0 views
arXiv:2611.17654v1 Announce Type: new Abstract: Movable antenna (MA) arrays improve multiuser throughput by repositioning their elements, but mechanical movement reduces transmission time and consumes actuation energy. Changes in user configurations create nonuniform demand for repositioning over time, while each movement determines which positions are reachable in subsequent slots. Thus, online control must decide when to spend a limited actuation-energy budget, not only where to position the antennas. We formulate causal effective-throughput maximization under displacement-dependent movement delay, inter-slot reachability, and a long-term average actuation-energy constraint. We propose Lyapunov-based online movable-antenna repositioning (LOMA), which adjusts the price of movement through a virtual queue and solves the resulting nonconvex position problem via proximal minorization--maximization, accounting for nonsmooth movement delay and promoting sparse antenna updates. We establish monotone ascent of the smoothed per-slot objective, a finite-horizon sample-path energy bound, almost-sure long-term budget feasibility under a Gaussian fading model with bounded large-scale gains, and a state-matched throughput comparison separating temporal energy allocation from per-slot optimization error. Simulations show that LOMA concentrates actuation after intermittent user changes and improves effective throughput over optimized fixed arrays and a per-slot-budget counterpart, while continuous channel drift favors more frequent incremental repositioning. Additional simulations with pilot-based channel estimates in sparse millimeter-wave channels show that LOMA retains most of the throughput benefits of online repositioning while operating near its prescribed long-term actuation-energy budget.
Source:https://arxiv.org/abs/2611.17654PreviousAn Axiomatic Framework for Reinforcement Learning in Open Multi-Agent SystemsNextBrain-inspired Beam Selection for mmWave RIS Communication
Related posts
Your Manager's Manager Has Booked a Skip Level Meeting With You
0dev.to

How to Make a Character Reference Sheet with AI (Prompts Included)
0dev.to Design
How Virtual Assistants Help Businesses Manage Their Work More Efficiently
0dev.to Productivity
DiffusionSSD: Achieving Extreme Visual Compression with Dual-stage Diffusion and Cross-level Near-Data Processing
0arXiv eess.IV
D2ToF: One-Step Diffusion for Depth Super-Resolution from Lightweight Single-Photon LiDAR Waveforms
0arXiv eess.IV
Few-Shot Open-Set Hyperspectral Pixel-Wise Classification with Spatial Aware Label Propagation
0arXiv eess.IV