RAVF: An Analysis and Verification Framework for Resource Allocation in Industrial Edge Computing with Probabilistic Model Checking and Machine Learning

Abstract Industrial edge computing has emerged as a pivotal paradigm to enhance the efficiency and responsiveness of industrial processes through decentralized data processing. Subsequently, ensuring the reliability of resource allocation becomes paramount. This paper presents a novel approach that combines probabilistic model checking and machine learning techniques for conducting reliability analysis of resource allocation in industrial edge computing environments. By leveraging probabilistic model checking, we can formally verify the system's behavior and assess the likelihood of resource allocation failures. Moreover, we integrate machine learning algorithms to predict potential resource allocation issues based on historical data patterns. This hybrid approach named RAVF framework offers a comprehensive perspective on reliability assessment, encompassing both theoretical analysis and data-driven insights. Through experiments and case studies, we demonstrate the effectiveness of our approach in identifying vulnerabilities in resource allocation and providing actionable insights for system optimization. The synergy between probabilistic model checking and machine learning equips industrial edge computing with a robust resource allocation framework that enhances overall system dependability.