The Internet of Things (IoT) connects billions of devices across domains such as transportation, healthcare, agriculture, and energy-creating highly dynamic, distributed, and heterogeneous environments. These characteristics pose significant challenges for control, coordination, scalability, and adaptability. In response, Multi-Agent Deep Reinforcement Learning (MADRL) has emerged as a promising paradigm by combining the decision-making intelligence of reinforcement learning with the collaborative capabilities of multi-agent systems. To leverage MADRL in building intelligent, resilient, and adaptive IoT systems, this review systematically explores the application of MADRL in IoT, categorizing contributions by application domains, learning architectures, and coordination strategies. We analyze how MADRL enables scalable resource allocation, routing optimization, energy efficiency, and fault detection in complex IoT ecosystems. Furthermore, we highlight key challenges-including scalability, non-stationarity, partial observability, and communication overhead-and discuss emerging solutions such as mean-field approximation, belief-state tracking, and federated MADRL.

To address key challenges in IoT systems, including efficient resource allocation, adaptive service composition, and Quality of Service (QoS) under dynamic conditions, we develop a framework called DRL-MAS integrating multi-agent systems (MAS) and deep reinforcement learning (DRL). DRL-MAS leverages MAS's decentralized decision-making capabilities and DRL's adaptive learning strengths to ensure scalability, energy efficiency, and responsiveness in distributed IoT systems. By incorporating edge computing, DRL-MAS minimizes dependency on centralized systems, reduces latency, and optimizes energy consumption. Experimental results demonstrate the DRL-MAS's effectiveness in dynamically optimizing service composition and resource management while complying with QoS requirements.

Nowadays, health care has become a constant concern of all countries around the world, especially after the emergence of the Coronavirus (COVID-19) and all its variants. Billions of dollars are being paid through the World Health Organization to improve health care. Scientific research laboratories play a pioneering role in this area as well. Due to the importance of information related to the patient and his medical history, it is necessary to exchange these information between various health centers in order to be better treated. Security in Health care information exchange (HIE) plays an important role because different healthcare facilities (HCFs) exchange sensitive data which can affect the patient’s privacy. Researchers propose many approaches in order to enhance security and maintain privacy. They also try to solve many drawbacks in this field like efficiency, accuracy, and scalability. Unfortunately, all the proposed techniques tackle some parameters and drop other ones. To decide which blockchain-based approach is more efficient for HIE systems, by evaluating each approach’s effectiveness based on its security, integrity, privacy, accuracy, scalability, efficiency, and latency qualities, we present a comparative analysis between a number of recent approaches in the HIE sector. Further, we use real patients’ data to measure each parameter, then we apply the Friedman test on the obtained results for each approach.

The advent of industry 4.0 (I4.0) has brought about significant advancements in manufacturing processes, leveraging advanced sensing and data analytics technologies to optimize efficiency. Within this paradigm, predictive maintenance (PdM) plays a crucial role in ensuring the reliability and availability of production systems. There are several existing approaches for PdM in I4.0, each with its own advantages and disadvantages. In this paper, we review the state-of-the-art related to PdM approaches in the context of I4.0. Our systematic literature review encompasses a comprehensive analysis of recent research, focusing on the different AI-based techniques employed in PdM applications. Through this survey, we aim to provide valuable insights into the current landscape of PdM methodologies and foster future innovations in this rapidly evolving field.

The advent of industry 4.0 (I4.0) has brought about significant advancements in manufacturing processes, leveraging advanced sensing and data analytics technologies to optimize efficiency. Within this paradigm, predictive maintenance (PdM) plays a crucial role in ensuring the reliability and availability of production systems. There are several existing approaches for PdM in I4.0, each with its own advantages and disadvantages. In this paper, we review the state-of-the-art related to PdM approaches in the context of I4.0. Our systematic literature review encompasses a comprehensive analysis of recent research, focusing on the different AI-based techniques employed in PdM applications. Through this survey, we aim to provide valuable insights into the current landscape of PdM methodologies and foster future innovations in this rapidly evolving field.

This article presents a novel approach to enhance asset management and resource sharing in intelligent industrial systems using Blockchain. We introduce a hybrid network architecture—termed the Hybrid Cyber-Physical System (HyCPS)—designed to facilitate decentralized and trustworthy data sharing across all layers. Central to this framework is a robust smart consensus protocol underpinned by a lightweight yet effective algorithm known as the Validation Trust Algorithm (VTA). Employing a combination of node-ranking and global decision-making strategies, the VTA ensures secure, transparent resource management. The integrated approach is implemented within the HyCPS architecture and rigorously validated through comprehensive simulations across diverse scenarios.

This article presents a novel approach to enhance asset management and resource sharing in intelligent industrial systems using Blockchain. We introduce a hybrid network architecture-termed the Hybrid Cyber-Physical System (HyCPS)-designed to facilitate decentralized and trustworthy data sharing across all layers. Central to this framework is a robust smart consensus protocol underpinned by a lightweight yet effective algorithm known as the Validation Trust Algorithm (VTA). Employing a combination of node-ranking and global decision-making strategies, the VTA ensures secure, transparent resource management. The integrated approach is implemented within the HyCPS architecture and rigorously validated through comprehensive simulations across diverse scenarios.

In the retail industry, Supply Chain Management (SCM) holds significant importance as it ensures the efficient movement of goods from suppliers to customers. In this intricate and fast-paced environment, the availability of accurate information and data is crucial. The purpose of this paper is to develop a framework that enhances forecasting accuracy and efficiency in Supply Chain (SC) operations within the retail industry. By analyzing the latest research and advancements in the field, thispaper seeks to contribute valuable insights into the potential of deep learning for SCM. The ultimate goal is to provide retailers with a reliable tool that empowers them to make informed decisions based on accurate predictions, thereby optimizing their SC operations and better meeting customer demands in the dynamic retail landscape. Furthermore, DLSTM-SCM framework updates dynamically the deployed Long Short-Term Memory (LSTM) models to predict sales for the upcoming day sales, utilizing historical sales data and incorporating statistical features such as lagging and shifting to enhance forecasting precision. The efficacy of DLSTM-SCM is demonstrated through its performance on real benchmarks, where it yielded significant improvements compared to existing methods.

Industry and academics are interested in industrial cyber-physical systems (ICPS). Complexity makes it hard to grasp these systems design and functioning. By offering FedGA-ICPS, a federated learning framework based on genetic algorithms, we can solve ICPSś performance and decision support. To simulate the structure and behavior of these systems, we use ICPS. FedGA-ICPS investigates the performance of the ICPS sensors by offering locally integrated learning models. Genetic algorithm then speeds up and improves federated learning aggregation. Transfer learning is used to disseminate model parameters across restricted entities. Fashion MNIST’s initiative achieved notable outcomes.

Industrial cyber-physical systems (ICPS) are heterogeneous inter-operating parts that can be physical, technical, networking, and even social like agent operators. Incrementally, they perform a central role in critical and industrial infrastructures, governmental, and personal daily life. Especially with the Industry 4.0 revolution, they became more dependent on the connectivity by supporting novel communication and distance control functionalities, which expand their attack surfaces that result in a high risk for cyber-attacks. Furthermore, regarding physical and social constraints, they may push up new classes of security breaches that might result in serious economic damages. Thus, designing a secure ICPS is a complex task, since this needs to guarantee security and harmonize the functionalities between the various parts that interact with different technologies. This article highlights the significance of cyber-security infrastructure and shows how to evaluate, prevent, and mitigate ICPS-based cyber-attacks. We carried out this objective by establishing an adequate semantics for ICPS’s entities and their composition, which includes social actors that act differently than mobile robots and automated processes. This article also provides the feasible attacks generated by a reinforcement learning mechanism based on multiple criteria that selects both appropriate actions for each ICPS component and the possible countermeasures for mitigation. To efficiently analyze ICPS’s security, we proposed a model-checking-based framework that relies on a set of predefined attacks from where the security requirements are used to assess how well the model is secure. Finally, to show the effectiveness of the proposed solution, we model, analyze, and evaluate the ICPS security on two real use cases.

Cyber-Physical Systems (CPS) have gained considerable interest in the last decade from both industry and academia. Such systems have proven particularly complex and provide considerable challenges to master their design and ensure their functionalities. In this paper, we intend to tackle some of these challenges related to the security and the resilience of CPS at the design level. We initiate a CPS modeling approach to specify such systems structure and behaviors, analyze their inherent properties and to overcome threats in terms of security and correctness. In this initiative, we consider a CPS as a network of entities that communicate through physical and logical channels, and which purpose is to achieve a set of tasks expressed as an ordered tree. Our modeling approach proposes a combination of the Coq theorem prover and the Maude rewriting system to ensure the soundness and correctness of CPS design. The introduced solution is illustrated through an automobile manufacturing case study.

Recent research focus more and more on IoT systems and their applications in order to make people life easier and controllable. The main aim is to expand IoT applications and services into various domains while ensuring communication and automated exchange between them. Recent research handles many issues related to IoT especially implementation, modeling, and deployment. However, many challenges need more deep and thorough analysis especially in terms of flexible modeling, extensible implementation, with respect to the privacy issue. This work focuses principally on modeling IoT systems dedicated to smart healthcare case. We attempt to address the emergency service by initiating a modeling mechanism for Healthcare Management System (HMS) by using UML diagrams, and propose an appropriate access control in order to reinforce it. Then, we ensure the correctness of the developed HMS by relying on the verification and validation based on a formal analysis that showed significant results by using Alloy tool.

The increasing proliferation of Cyber-Physical Systems (CPS) in industry and academia bring researchers to work on CPS architectures and models seeking for enhancing the performance and gaining the full potential of the system. This work aims to provide a holistic view of the initiatives done during the last five years in modeling and designing CPS. We define three major classes of CPS covering Smart Healthcare, Smart Manufacturing, and Smart City. We also provide a review of the recent developed architectures and models for each class. Based on the surveyed literature, we identified many open issues and we suggested possible future research directions.

One of the main challenges in the development process of secure systems is how to detect as early as possible the system's vulnerabilities and weaknesses, and also how to quantify the severity of attacks through them. In this paper, we rely on the concept of attack surfaces to implement a secure cyber physical system in Java. Attack surfaces can be sometimes detected automatically, regarding the used language, by matching them against known attacks still is a step apart. Further, systems and attacks are not usually modeled with compatible formalism. This paper develops a modeling framework that automates the whole process by generating attacks for cyber physical systems. First, we formalize a system using UML class and activity diagrams. Further, we use UML to develop a meta language for cyber physical systems, cyber attacks, and cyber counter measures. The framework instantiates the dependent-application diagrams for the domain/application in test, searches for the existing attack surfaces; then it generates the possible attacks that might exploit the found vulnerabilities/weaknesses. Further the proposed framework generates the proper java code for the composition counter measures, attacks, and CPS models.

SysML activity diagrams are OMG/INCOSE standard used for modeling and analyzing probabilistic systems. In this paper, we propose a formal verification framework that is based on PRISM probabilistic symbolic model checker to verify the correctness of these diagrams. To this end, we present an efficient algorithm that transforms a composition of SysML activity diagrams to an equivalent probabilistic automata encoded in PRISM input language. To clarify the quality of our verification framework, we formalize both SysML activity diagrams and PRISM input language. Finally, we demonstrate the effectiveness of our approach by presenting a case study.

The standard OMG/INCOSE SysML activity diagrams are behavioral models for specifying and analyzing probabilistic systems. In this paper, we present a formal verification framework for these diagrams that helps to mitigate the state-explosion problem in probabilistic model checking. To do so, we propose to reduce the size of SysML activity diagrams by eliminating and merging precise behaviors. The resulting model is checked using Probabilistic Computation Tree Logic (PCTL) properties. Moreover, we present a calculus for SysML activity diagrams (NuAC) that captures their underlying semantics. In addition, we prove the soundness of our approach by defining a probabilistic weak simulation relation between the semantics of the abstract and the concrete models. This relation is shown to preserve the satisfaction of the PCTL properties. Finally, we demonstrate the effectiveness of our approach on an online shopping system case study.

In this paper, we address the issue of security verification and evaluation of systems at the design level. To this end, we elaborate a practical and formal framework that enables security risk assessment and security requirements verification on systems that are designed using SysML activity diagrams. Our approach is based on probabilistic adversarial interactions between potential attackers and the system design models. These interactions result in a global model that is used to quantify security risks by applying probabilistic model-checking. We rely on a standard catalogue of attack patterns to build a library of attacks' design patterns. To demonstrate the effectiveness of our approach, we apply it on a real-life case study related to the Secure Real Time Streaming Protocol.