Explainable AI for Radar Resource Management: Modified LIME in Deep Reinforcement Learning

2025-06-25

AI Summary

Deep reinforcement learning (DRL) achieves superior decision-making performance in radar resource management (RRM), yet its black-box nature severely undermines interpretability. Existing eXplainable AI (XAI) methods—such as LIME—suffer from low local fidelity due to their failure to account for feature correlations. To address this, we propose DL-LIME: a novel XAI framework that integrates deep learning into LIME’s sampling process to construct a feature-correlation-aware, adaptive perturbation mechanism, enabling high-fidelity local explanations tailored to RRM decision contexts. Experimental results demonstrate that DL-LIME significantly outperforms standard LIME in both explanation fidelity and task performance. Moreover, it successfully identifies critical state variables influencing resource scheduling—including target density, signal-to-noise ratio, and mission priority—thereby providing both rigorous interpretability guarantees and a practical framework for trustworthy intelligent radar scheduling.

Technology Category

• Machine Learning: Transparent, Interpretable, Explainable ML
• Humans and AI: Explainable AI (XAI) for Human Understanding
• Knowledge Representation and Reasoning: Description Logics

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Deep reinforcement learning has been extensively studied in decision-making processes and has demonstrated superior performance over conventional approaches in various fields, including radar resource management (RRM). However, a notable limitation of neural networks is their ``black box" nature and recent research work has increasingly focused on explainable AI (XAI) techniques to describe the rationale behind neural network decisions. One promising XAI method is local interpretable model-agnostic explanations (LIME). However, the sampling process in LIME ignores the correlations between features. In this paper, we propose a modified LIME approach that integrates deep learning (DL) into the sampling process, which we refer to as DL-LIME. We employ DL-LIME within deep reinforcement learning for radar resource management. Numerical results show that DL-LIME outperforms conventional LIME in terms of both fidelity and task performance, demonstrating superior performance with both metrics. DL-LIME also provides insights on which factors are more important in decision making for radar resource management.

Problem

• Enhancing explainability in radar resource management AI
• Addressing feature correlation neglect in LIME method
• Improving decision transparency in deep reinforcement learning

Innovation

• Modified LIME for explainable AI
• Deep learning integrated sampling process
• DL-LIME enhances fidelity and performance