Explaining Strategic Decisions in Multi-Agent Reinforcement Learning for Aerial Combat Tactics
NATO Journal of Science and Technology · 10.03.2026
DOI: 10.14339/sto-msg-217-05 ↗Kurz gefasst
Fachartikel in der Zeitschrift NATO Journal of Science and Technology, online erschienen am 10.03.2026. Verfasst von Ardian Selmonaj und 4 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
Artificial intelligence (AI) is pivotal in shaping the future technological landscape. Multi-Agent Reinforcement Learning (MARL) has emerged as a significant AI technology for simulating complex dynamics across various domains, enabling novel potentials for advanced strategic planning and coordination among autonomous agents. However, its practical deployment in sensitive military contexts is constrained by the lack of explainability: a critical factor for reliability, safety, strategic validation, and human-machine interaction. This paper reviews the latest advancements in explainability within MARL and presents novel use cases, emphasizing its indispensability for examining agents’ decision-making processes. Existing techniques are critically assessed and associated with the domain of military strategies, focusing on simulated air combat scenarios. The concept of a novel information-theoretic explainability descriptor to analyse the cooperation capabilities of agents is then introduced. The aim of this research is to highlight the necessity of precisely understanding AI decisions and aligning these artificially generated tactics with human understanding and strategic military doctrines, thereby enhancing the transparency and reliability of AI systems. By illuminating the crucial importance of explainability in advancing MARL for operational defence, this work supports not only strategic planning but also the training of military personnel with insightful and comprehensible analyses.
Zitieren
APA
Selmonaj, A., Antonucci, A., Schneider, A., Sommer, M., & Rüegsegger, M. (2026). Explaining Strategic Decisions in Multi-Agent Reinforcement Learning for Aerial Combat Tactics. NATO Journal of Science and Technology. https://doi.org/10.14339/sto-msg-217-05
Vancouver
Selmonaj A, Antonucci A, Schneider A, Sommer M, Rüegsegger M. Explaining Strategic Decisions in Multi-Agent Reinforcement Learning for Aerial Combat Tactics. NATO Journal of Science and Technology. 2026. doi:10.14339/sto-msg-217-05
BibTeX
@article{Selmonaj2026,
author = {Selmonaj, Ardian and Antonucci, Alessandro and Schneider, Adrian and Sommer, Matthias and Rüegsegger, Michael},
title = {Explaining Strategic Decisions in Multi-Agent Reinforcement Learning for Aerial Combat Tactics},
journal = {NATO Journal of Science and Technology},
year = {2026},
doi = {10.14339/sto-msg-217-05}
}
EUVIMED