The editorial summarises the special issue that is based on contributions from three international conferences focused on mathematics, informatics, data-driven analytics, intelligent systems, and information technology. This special issue consists of seven articles.
Background: Machine learning-based intrusion detection systems are being used increasingly to keep internet of things (IoT) environments safe, but the fact that they can be tweaked by adversaries limits their dependability, especially in resource-constrained, real-time deployments.Objective: This study proposes and evaluates an adversarially robust intrusion detection framework, integrated into the OneM2M service layer, for real-time edge IoT security, where the goal is to keep communications safe, even when attackers act in crafty ways.Methods: Within the OneM2M analytics engine, the framework combines statistical anomaly filtering, feature squeezing, adversarial training, confidence driven rejection and federated learning. It is tested across white-box, black-box, transfer learning style and even data poisoning situations using IoT security datasets, OneM2M traffic patterns and edge-centred deployment setups.Results: This layered defence approach improves resilience against adversarial attacks while keeping detection performance, inference efficiency and resource use within defined limits. In the evaluation part, it shows operational feasibility too, even across varying traffic loads and different edge deployment configurations.Conclusion: Integrating a multilayer adversarial-defence pipeline into standardized OneM2M middleware can help reinforce real-time IoT intrusion detection while not completely giving up edge-resource limits. It is more robust, workable in practice and less easy for attackers to slip through in the usual environment.
Background: In the field of distributed artificial intelligence, multi-agent systems (MAS) collaborate to resolve complex problems. To operate efficiently, these systems require effective coordination mechanisms. Metaheuristic algorithms are now standard for decentralized optimization and agent task distribution.Objective: This study evaluates the strengths and limitations of common metaheuristics used for MAS path planning and task coordination. Additionally, it introduces comparative simulation studies to evaluate the performance of baseline and dynamic metaheuristic methods specifically in MAS path planning.Methods: This paper presents a comprehensive narrative review of metaheuristic approaches for MAS coordination to identify current research gaps. It evaluates recent literature to identify their strengths, weaknesses and future trends. For practical validation, this review is coupled with a comparative simulation study using NetLogo to evaluate 50 agents in a 2D grid. The study specifically compares the standard crow search algorithm (SCSA) and standard particle swarm optimization (PSO) against their adaptive and modified versions (AAP-CSA, DEC-DAPCSA, MCSA, BPSO and LDPSO). This dual approach effectively bridges the gap between theoretical overview and practical performance evaluation.Results: The findings indicate that metaheuristic algorithms significantly affect MAS performance. Specifically, standard algorithms (SCSA and PSO) often trap in local optima, but their dynamic variants improve both fitness and convergence speed.Conclusion: The evaluated dynamic parameter strategies outperformed standard versions in both solution quality and search time. Ultimately, selecting the right metaheuristic and applying dynamic parameter control are essential for solving complex MAS coordination problems.
Background: In recent years, the Internet of Medical Things (IoMT) has transformed the healthcare sector through real-time patient monitoring and continuous data collection. However, transmitting sensitive medical information over public networks exposes IoMT systems to significant security threats, while emerging quantum computing technologies challenge the reliability of traditional cryptographic systems.Objective: The objective of this study is to propose PQAC-BIoMT, a secure and robust model for remote user authentication and access control in IoMT environments, capable of withstanding both conventional and quantum attacks.Methods: This article proposes a decentralized authentication framework that integrates post-quantum cryptography using Kyber Public-Key Encryption (Kyber-PKE) into blockchain-based smart contracts. Fog computing nodes are used to reduce the authentication latency and improve the system scalability. A role-based authorization mechanism is integrated to link user identities to functional roles and enforce authorization to medical data and system resource access. Formal security verification is conducted using Burrows-Abadi-Needham (BAN) logic to validate the correctness of authentication, and the Automated Validation of Internet Security Protocols and Applications (AVISPA) tool is used to assess resistance to known attacks. PQAC-BIoMT is further evaluated through a comparative analysis of the computational load, energy consumption and security properties.Results: Our security analysis demonstrates that PQAC-BIoMT effectively resists common attacks while providing quantum-resistant protection against them. The performance evaluation shows that the proposed scheme achieves relatively lower computational and energy overhead compared to existing approaches, making it suitable for resource-constrained IoMT devices.Conclusion: The proposed PQAC-BIoMT scheme delivers a secure, quantum-resilient authentication and authorization mechanism for IoMT systems, enhancing both data protection and operational efficiency, which can support practical deployment in real-world IoMT applications.
Background: The convergence of large language models (LLMs) and gamification has emerged as a promising direction for enhancing user engagement, personalisation and interaction quality across digital systems. While prior studies have largely examined LLMs or gamification in isolation, limited work has systematically explored their combined role in redefining engagement paradigms beyond static reward-based mechanisms.Objective: This review systematically examines how LLMs are integrated into gamified systems, identifies their application contexts and functional roles and synthesises reported benefits, challenges and methodological trends across existing empirical studies.Methods: This systematic literature review examines peer-reviewed studies published between 2020 and 2025, following established systematic review protocols and the PRISMA 2020 framework. A total of 718 articles were retrieved from Scopus, IEEE Xplore, ScienceDirect and the ACM Digital Library. After applying predefined inclusion and exclusion criteria, 27 studies were selected for in-depth analysis. A quality assessment process was conducted to evaluate methodological rigour, consistency and reporting clarity, ensuring the reliability of the synthesis.Results: The review identifies key application domains where LLMs are integrated into gamified systems, including intelligent non-player characters, adaptive educational feedback, persuasive gamified environments and embodied digital avatars for social and behavioural simulations. Across these contexts, LLMs were found to substantially enhance personalisation, narrative immersion and learner motivation, complementing the motivational structures of gamification. However, recurring challenges were also reported, including hallucination risks, ethical concerns in social and extended reality contexts and misalignment between LLM-driven system behaviour and user expectations.Conclusion: This review provides a comprehensive synthesis of the opportunities, limitations and methodological trends associated with integrating LLMs into gamified environments. By consolidating evidence across diverse application domains, it offers practical and theoretical guidance for researchers, educators and system designers on responsible and effective use of LLMs to support adaptive, engaging and ethically grounded gamified systems.
