@article{ART003300287},
author={Woo-Hyeon Moon and Seo-Ho Lee and Won-Seok Jang and Ji-Seok Yoon and Hyeon-Mo Kim and Ju-Mi Park and Jae-Bok Sung},
title={Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards},
journal={Journal of The Korea Society of Computer and Information},
issn={1598-849X},
year={2026},
volume={31},
number={1},
pages={261-270}
TY - JOUR
AU - Woo-Hyeon Moon
AU - Seo-Ho Lee
AU - Won-Seok Jang
AU - Ji-Seok Yoon
AU - Hyeon-Mo Kim
AU - Ju-Mi Park
AU - Jae-Bok Sung
TI - Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards
JO - Journal of The Korea Society of Computer and Information
PY - 2026
VL - 31
IS - 1
PB - The Korean Society Of Computer And Information
SP - 261
EP - 270
SN - 1598-849X
AB - In this paper, we propose a reinforcement learning-based decision-making algorithm tailored for multi-weapon and multi-threat combat environments. The proposed method separates actor-critic structures by weapon type and incorporates self-attention and cross-attention mechanisms to enable interaction-aware learning across agents. A responsibility-based reward function is designed to evaluate the contribution of each weapon to combat outcomes, promoting cooperative behavior and preventing policy bias. The algorithm is implemented using the Stable-Baselines3 library and validated through tactical simulations. Experimental results demonstrate that our method achieves superior strategic efficiency and collaboration performance compared to conventional PPO-based models, especially under sparse reward conditions and resource constraints.
KW - Reinforcement Learning;Multi-Agent;Threat Assessment;Weapon Allocation;;Attention Mechanism;Simulation
DO -
UR -
ER -
Woo-Hyeon Moon, Seo-Ho Lee, Won-Seok Jang, Ji-Seok Yoon, Hyeon-Mo Kim, Ju-Mi Park and Jae-Bok Sung. (2026). Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards. Journal of The Korea Society of Computer and Information, 31(1), 261-270.
Woo-Hyeon Moon, Seo-Ho Lee, Won-Seok Jang, Ji-Seok Yoon, Hyeon-Mo Kim, Ju-Mi Park and Jae-Bok Sung. 2026, "Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards", Journal of The Korea Society of Computer and Information, vol.31, no.1 pp.261-270.
Woo-Hyeon Moon, Seo-Ho Lee, Won-Seok Jang, Ji-Seok Yoon, Hyeon-Mo Kim, Ju-Mi Park, Jae-Bok Sung "Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards" Journal of The Korea Society of Computer and Information 31.1 pp.261-270 (2026) : 261.
Woo-Hyeon Moon, Seo-Ho Lee, Won-Seok Jang, Ji-Seok Yoon, Hyeon-Mo Kim, Ju-Mi Park, Jae-Bok Sung. Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards. 2026; 31(1), 261-270.
Woo-Hyeon Moon, Seo-Ho Lee, Won-Seok Jang, Ji-Seok Yoon, Hyeon-Mo Kim, Ju-Mi Park and Jae-Bok Sung. "Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards" Journal of The Korea Society of Computer and Information 31, no.1 (2026) : 261-270.
Woo-Hyeon Moon; Seo-Ho Lee; Won-Seok Jang; Ji-Seok Yoon; Hyeon-Mo Kim; Ju-Mi Park; Jae-Bok Sung. Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards. Journal of The Korea Society of Computer and Information, 31(1), 261-270.
Woo-Hyeon Moon; Seo-Ho Lee; Won-Seok Jang; Ji-Seok Yoon; Hyeon-Mo Kim; Ju-Mi Park; Jae-Bok Sung. Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards. Journal of The Korea Society of Computer and Information. 2026; 31(1) 261-270.
Woo-Hyeon Moon, Seo-Ho Lee, Won-Seok Jang, Ji-Seok Yoon, Hyeon-Mo Kim, Ju-Mi Park, Jae-Bok Sung. Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards. 2026; 31(1), 261-270.
Woo-Hyeon Moon, Seo-Ho Lee, Won-Seok Jang, Ji-Seok Yoon, Hyeon-Mo Kim, Ju-Mi Park and Jae-Bok Sung. "Reinforcement Learning-Based Multi-Target Threat Assessment and Weapon Assignment Algorithm Using Attention and Responsibility-Based Rewards" Journal of The Korea Society of Computer and Information 31, no.1 (2026) : 261-270.
