Abstract:Public safety agencies have been working on the modernization of their communication networks and the enhancement of their mission-critical capabilities with novel technologies and applications. As part of these efforts, migrating from traditional land mobile radio (LMR) systems toward cellular-enabled, next-generation, mission-critical networks is at the top of these agencies’ agendas. In this paper, we provide an overview of cellular technologies ratified by the 3rd Generation Partnership Project (3GPP) to e… Show more
“…While some works, such as (Renda et al, 2021;Marcu et al, 2023;Barnard et al, 2022) aim to understand the decisions of ML models, others, such as (Terra et al, 2020;Morichetta et al, 2019;Spyrou and Kappatos, 2023) also uses XAI methods to explain relationships between network variables, either for cognitive purposes or for selecting relevant attributes when training ML models. Additionally, XAI constitutes a fundamental tool for regulatory compliance by facilitating traceability and explanation of AI decisions, something essential in critical applications such as healthcare, autonomous transportation, and cybersecurity management in these next-generation networks (Saafi et al, 2021). For example, the European Union's General Data Protection Regulation (GDPR) requires that AI models that make decisions about individuals be explainable (Bayamlıoglu, 2022).…”
As the 5G era progresses and the research community shifts its focus to the future 6G era, an unprecedented surge in the adoption of Artificial Intelligence (AI) techniques for network development and operation is expected. AI is envisioned to play a crucial role in 6G networks, enabling intelligent network management, enhanced user experience, higher security, and unprecedented levels of connectivity. However, the opaque nature of Machine Learning (ML) models has prompted a shift towards Explainable AI (XAI) techniques to enhance decision-making transparency and auditability. Despite the promises of XAI, computational costs remain a critical consideration. This study investigates the temporal and energy costs associated with four prominent XAI techniques: SHapley Additive exPlanations (SHAP), Local Interpretable Model-agnostic Explanations (LIME), Permutation Importance (PI), and Morris Sensitivity (MS). These techniques are applied to four ML models in two distinct 5G network scenarios. Our results show that MS emerged as the most time-efficient and energy-conserving XAI method, demonstrating consistent feature relevance across various ML models and datasets, affirming its efficacy in explaining model decisions.
“…While some works, such as (Renda et al, 2021;Marcu et al, 2023;Barnard et al, 2022) aim to understand the decisions of ML models, others, such as (Terra et al, 2020;Morichetta et al, 2019;Spyrou and Kappatos, 2023) also uses XAI methods to explain relationships between network variables, either for cognitive purposes or for selecting relevant attributes when training ML models. Additionally, XAI constitutes a fundamental tool for regulatory compliance by facilitating traceability and explanation of AI decisions, something essential in critical applications such as healthcare, autonomous transportation, and cybersecurity management in these next-generation networks (Saafi et al, 2021). For example, the European Union's General Data Protection Regulation (GDPR) requires that AI models that make decisions about individuals be explainable (Bayamlıoglu, 2022).…”
As the 5G era progresses and the research community shifts its focus to the future 6G era, an unprecedented surge in the adoption of Artificial Intelligence (AI) techniques for network development and operation is expected. AI is envisioned to play a crucial role in 6G networks, enabling intelligent network management, enhanced user experience, higher security, and unprecedented levels of connectivity. However, the opaque nature of Machine Learning (ML) models has prompted a shift towards Explainable AI (XAI) techniques to enhance decision-making transparency and auditability. Despite the promises of XAI, computational costs remain a critical consideration. This study investigates the temporal and energy costs associated with four prominent XAI techniques: SHapley Additive exPlanations (SHAP), Local Interpretable Model-agnostic Explanations (LIME), Permutation Importance (PI), and Morris Sensitivity (MS). These techniques are applied to four ML models in two distinct 5G network scenarios. Our results show that MS emerged as the most time-efficient and energy-conserving XAI method, demonstrating consistent feature relevance across various ML models and datasets, affirming its efficacy in explaining model decisions.
“…On the application side, we observe that public safety organizations have already begun to shift from traditional land mobile radio to cellular communications systems, leveraging a new set of deployed devices to meet mission-critical requirements and target new public-safety broadband applications. Accordingly, 3GPP Rel-16 targets defining the common architecture for public safety and commercial ProSe services [98]. In the case of public safety, maintaining ProSe discovery and communication is especially critical when the UE resides outside the coverage area of the cellular network, e.g., in the case of disaster management in remote areas.…”
for their assistance at every stage of the 3 years of my Ph.D. I also would like to thank Simona Lohan and Alex Ometov for their guidance and help throughout the A-WEAR project. My gratitude extends to Claudia Campolo for her valuable feedback on the research questions, and to Sergey Andreev and Olga Galinina for their patience and time spent teaching me.I would like to offer my special thanks to Sara Pizzi (and to her family), to my friend and colleague in parallel, for her support and love in the real world, and for her insightful comments and suggestions in the work world.I could not have undertaken this journey without my sister Olga Chukhno. I am thankful for her unwavering support, positive thinking, encouragement, and love. Special thanks here.Many thanks to my Italian family, which I obtained, and I am sure, I will never lose
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.