Purpose: This paper focuses on examining the robustness of a cooperative Multi-Agent Reinforcement Learning (MARL)-based Intrusion Detection System (IDS) for intrusion detection in decentralised 5G security settings. Even though MARL techniques have proven effective against dynamic threats in decentralized 5G networks, current research has not considered any adversarial scenarios at all.
Methods: A cooperative MARL-based Intrusion Detection System was developed through the CRISP-DM approach. Radio Access Network (RAN), MEC, and Core agents were trained using Centralised Training with Decentralised Execution (CTDE) and Deep Q-Network (DQN) methods. The algorithm was tested on the NSL-KDD and UNSW-NB15 datasets against Fast Gradient Sign Method (FGSM) evasion attacks (ε = 0.05-0.30) and Byzantine poisoning attacks with 5%, 10%, and 20% compromised agents.
Result: The model achieved 96.94% accuracy on NSL-KDD and 85.15% on UNSW-NB15 in clean scenarios. The FGSM attack at ε = 0.20 resulted in substantial performance deterioration, leading to accuracy drops of 50.14 and 45.26 percentage points, respectively, and a simultaneous increase in false positives. Byzantine poisoning produced smaller but persistent decreases in accuracy of 12.03 and 2.62 percentage points, respectively.
Novelty: This study provides among the first empirical evaluations of adversarial fragility in cooperative MARL-based intrusion detection within distributed 5G-oriented security abstractions, demonstrating that cooperative intelligence alone does not guarantee adversarial robustness.
B. Ndlovu, Kudzaishe Lawal Chizengwe· Scientific Journal of Inform...· 0 citations
Purpose: Achieving global food security while maintaining environmentally sustainable agricultural systems remains a critical challenge amid population growth, climate variability, and resource constraints. Artificial Intelligence (AI) and the Internet of Things (IoT) have emerged as transformative technologies that support data-driven agricultural practices. This study systematically examines the applications, opportunities, challenges, and adoption factors of AI-IoT integration in smart agriculture, with particular emphasis on its potential contributions to food security and sustainable farming.
Methods: A systematic literature review (SLR) was conducted following the PRISMA 2020 guidelines. Publications were retrieved from Scopus, IEEE Xplore, and ScienceDirect covering the period 2020-2024. From an initial 431 records, 17 empirical studies met the inclusion criteria and were analysed using narrative and thematic synthesis.
Result: The review shows that AI-IoT technologies are primarily applied in crop disease detection, precision agriculture, environmental monitoring, yield prediction, and livestock health monitoring. These technologies enable real-time decision support, early disease detection, productivity improvement, and resource optimisation. However, challenges remain, including data limitations, infrastructure constraints, integration complexity, and high deployment costs.
Novelty: The study proposes a layered smart agriculture framework linking technological infrastructure, application domains, adoption conditions, operational outcomes, and sustainability impacts. The findings highlight key factors necessary for successful implementation, including infrastructure readiness, affordability, technological reliability, and capacity development for farmers. Crucially, the review demonstrates that the field of AI-IoT smart agriculture is technically advanced but socio-technically incomplete: the evidence base is dominated by proof-of-concept studies from Asia, with no empirical representation from Africa or the Americas, creating what this study terms an AI-IoT agricultural equity gap that fundamentally limits the technology’s contribution to global food security. The five-layered framework introduced here provides the first inductively derived organising structure that explicitly connects AI-IoT infrastructure to SDG-aligned food security outcomes, offering a replicable analytical scaffold for future empirical and policy research in this domain.
B. Ndlovu, Kudakwashe Maguraushe· Scientific Journal of Inform...· 0 citations