SparsePilot: Belief-Guided Network Planning under Sparse Wireless Measurements

2026-08-10

AI Summary

This work proposes SparsePilot, a framework that jointly optimizes active sensing and coverage control for unmanned aerial vehicles under the constraint of sparse wireless signal strength measurements. The approach formulates spatial exploration as a multi-armed bandit problem, employing an upper confidence bound strategy to select information-rich regions. It constructs a coverage belief map as the state input to a deep reinforcement learning controller, which generates continuous motion actions. SparsePilot achieves the first end-to-end joint optimization of exploration and control under sparse feedback by replacing full observations with belief states, and establishes a theoretical connection among sparse probing, belief estimation error, and performance gap. Experiments across seven urban digital twins demonstrate that SparsePilot significantly outperforms baseline methods using only approximately 3.1% of the full observation budget and exhibits strong generalization to unseen wireless environments.

Abstract

Unmanned aerial vehicles (UAVs) have emerged as a promising solution for on-demand wireless coverage planning in urban environments. Existing learning-based UAV control methods, however, typically rely on continuous access to dense user-level received signal strength (RSS) measurements. Such full-observation assumptions are difficult to satisfy in real-world deployments due to the high cost and limited availability of dense wireless feedback. Sparse-feedback decision making under severe observation constraints therefore represents a fundamental challenge. To fill this gap, we propose SparsePilot, a measurement-efficient sensing-control framework that couples active wireless probing with belief-guided network control. SparsePilot formulates spatial probing as a multi-armed bandit problem over grid cells, uses upper confidence bound probing to select informative regions, and aggregates sparse RSS measurements into a coverage belief map. A deep reinforcement learning controller then uses this belief state to generate continuous UAV mobility actions, while the full wireless state remains hidden from the policy. We further provide a theoretical analysis connecting sparse probing, belief estimation error, and the sparse-feedback performance gap. Experiments across seven urban digital twins show that SparsePilot achieves superior coverage restoration performance while using only about 3.1% of the full-observation measurement budget and demonstrates strong cross-scene generalization to unseen urban-scale wireless environments.

Problem

• sparse wireless measurements
• UAV coverage planning
• observation constraints
• wireless feedback
• belief-guided control

Innovation

• sparse feedback
• belief-guided control
• active probing
• multi-armed bandit
• deep reinforcement learning