A Spatially Informed Gaussian Process UCB Method for Decentralized Coverage Control

2025-11-04

AI Summary

This paper addresses the decentralized coverage control problem for multi-agent systems in unknown spatial environments. We propose a Gaussian process (GP)-driven method that jointly balances exploration and exploitation. Our key contribution is the first integration of the GP-Upper Confidence Bound (GP-UCB) strategy into a decentralized framework, coupled with an inducing-point-based greedy update mechanism to enable scalable online GP learning. Each agent operates solely on local observations and neighbor communications, performing distributed optimization to minimize its local cost function. The approach maintains high modeling fidelity while significantly improving exploration efficiency. Simulation results demonstrate approximately 40% faster convergence and over 25% higher final coverage compared to baseline methods, effectively reconciling uncertainty-aware exploration with task-oriented coverage performance.

Technology Category

• Reasoning under Uncertainty: Sequential Decision Making
• Multiagent Systems: Multiagent Systems under Uncertainty
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We present a novel decentralized algorithm for coverage control in unknown spatial environments modeled by Gaussian Processes (GPs). To trade-off between exploration and exploitation, each agent autonomously determines its trajectory by minimizing a local cost function. Inspired by the GP-UCB (Upper Confidence Bound for GPs) acquisition function, the proposed cost combines the expected locational cost with a variance-based exploration term, guiding agents toward regions that are both high in predicted density and model uncertainty. Compared to previous work, our algorithm operates in a fully decentralized fashion, relying only on local observations and communication with neighboring agents. In particular, agents periodically update their inducing points using a greedy selection strategy, enabling scalable online GP updates. We demonstrate the effectiveness of our algorithm in simulation.

Problem

• Decentralized coverage control in unknown spatial environments
• Balancing exploration and exploitation using Gaussian Processes
• Scalable online updates with local observations and neighbor communication

Innovation

• Decentralized coverage control using Gaussian Processes
• Combines expected cost with variance-based exploration
• Scalable online GP updates with greedy selection