Learning Decentralized Strategies for a Perimeter Defense Game with Graph Neural Networks

—We consider the problem of finding decentralized strategies for multi-agent perimeter defense games. In this work, we design a graph neural network-based learning frame- work to learn a mapping from defenders’ local perceptions and the communication graph to defenders’ actions such that the learned actions are close to that generated by a centralized expert algorithm. We demonstrate that our proposed networks stay closer to the expert policy and are superior to other baseline algorithms by capturing more intruders. Our GNN- based networks are trained at a small scale and can generalize to large scales. To validate our results, we run perimeter defense games in scenarios with different team sizes and initial configurations to evaluate the performance of the learned networks.

[1]  N. Sundararajan,et al.  A Decentralized Multirobot Spatiotemporal Multitask Assignment Approach for Perimeter Defense , 2022, IEEE Transactions on Robotics.

[2]  Vijay Kumar,et al.  Robust Multi-Robot Active Target Tracking Against Sensing and Communication Attacks , 2021, IEEE Transactions on Robotics.

[3]  Efstathios Bakolas,et al.  Guarding a Convex Target Set From an Attacker in Euclidean Spaces , 2021, IEEE Control Systems Letters.

[4]  Vijay Kumar,et al.  Defending a Perimeter from a Ground Intruder Using an Aerial Defender: Theory and Practice , 2021, 2021 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR).

[5]  Vijay R. Kumar,et al.  Graph Neural Networks for Decentralized Multi-Robot Target Tracking , 2021, 2022 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR).

[6]  Meir Pachter,et al.  Guarding a Circular Target By Patrolling its Perimeter , 2020, 2020 59th IEEE Conference on Decision and Control (CDC).

[7]  Vijay Kumar,et al.  Perimeter-defense Game between Aerial Defender and Ground Intruder , 2020, 2020 59th IEEE Conference on Decision and Control (CDC).

[8]  Zhe Liu,et al.  Message-Aware Graph Attention Networks for Large-Scale Multi-Robot Path Planning , 2020, IEEE Robotics and Automation Letters.

[9]  George J. Pappas,et al.  Adaptive Partitioning for Coordinated Multi-agent Perimeter Defense , 2020, 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[10]  Fernando Gama,et al.  Graph Neural Networks: Architectures, Stability, and Transferability , 2020, Proceedings of the IEEE.

[11]  Matthew Gombolay,et al.  Learning Scheduling Policies for Multi-Robot Coordination With Graph Attention Networks , 2020, IEEE Robotics and Automation Letters.

[12]  Giuseppe Loianno,et al.  Experimental Evaluation and Characterization of Radioactive Source Effects on Robot Visual Localization and Mapping , 2020, IEEE Robotics and Automation Letters.

[13]  Xu Liu,et al.  SLOAM: Semantic Lidar Odometry and Mapping for Forest Inventory , 2019, IEEE Robotics and Automation Letters.

[14]  Natalia Gimelshein,et al.  PyTorch: An Imperative Style, High-Performance Deep Learning Library , 2019, NeurIPS.

[15]  F. Bullo,et al.  Matching-based capture strategies for 3D heterogeneous multiplayer reach-avoid differential games , 2019, Autom..

[16]  Giuseppe Loianno,et al.  MAVNet: An Effective Semantic Segmentation Micro-Network for MAV-Based Tasks , 2019, IEEE Robotics and Automation Letters.

[17]  Vijay Kumar,et al.  Learning Decentralized Controllers for Robot Swarms with Graph Neural Networks , 2019, CoRL.

[18]  Vijay Kumar,et al.  Decentralization of Multiagent Policies by Learning What to Communicate , 2019, 2019 International Conference on Robotics and Automation (ICRA).

[19]  Vijay Kumar,et al.  Local-game Decomposition for Multiplayer Perimeter-defense Problem , 2018, 2018 IEEE Conference on Decision and Control (CDC).

[20]  Pratap Tokekar,et al.  Resilient Active Target Tracking With Multiple Robots , 2018, IEEE Robotics and Automation Letters.

[21]  Antonio G. Marques,et al.  Convolutional Neural Network Architectures for Signals Supported on Graphs , 2018, IEEE Transactions on Signal Processing.

[22]  Pratap Tokekar,et al.  Sensor Assignment Algorithms to Improve Observability While Tracking Targets , 2017, IEEE Transactions on Robotics.

[23]  Pratap Tokekar,et al.  Active Target Tracking With Self-Triggered Communications in Multi-Robot Teams , 2017, IEEE Transactions on Automation Science and Engineering.

[24]  Dongsuk Kum,et al.  Drone-Assisted Disaster Management: Finding Victims via Infrared Camera and Lidar Sensor Fusion , 2016, 2016 3rd Asia-Pacific World Congress on Computer Science and Engineering (APWC on CSE).

[25]  Mehdi Bennis,et al.  UAV-Assisted Heterogeneous Networks for Capacity Enhancement , 2016, IEEE Communications Letters.

[26]  Mo Chen,et al.  Multiplayer reach-avoid games via low dimensional solutions and maximum matching , 2014, 2014 American Control Conference.

[27]  Vijay Kumar,et al.  Minimum snap trajectory generation and control for quadrotors , 2011, 2011 IEEE International Conference on Robotics and Automation.

[28]  Wolfram Burgard,et al.  A real-time algorithm for mobile robot mapping with applications to multi-robot and 3D mapping , 2000, Proceedings 2000 ICRA. Millennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings (Cat. No.00CH37065).

[29]  George J. Pappas,et al.  Optimal Multi-robot Perimeter Defense Using Flow Networks , 2021, DARS.

[30]  Vijay Kumar,et al.  A Review of Multi Agent Perimeter Defense Games , 2020, GameSec.

[31]  Morten Bisgaard,et al.  Adaptive Surveying and Early Treatment of Crops with a Team of Autonomous Vehicles , 2011, ECMR.

[32]  Jonathan M. Garibaldi,et al.  Multi-Robot Search and Rescue: A Potential Field Based Approach , 2007 .