Internet of Things (IoT) ecosystems refer to interconnected intelligent devices that enhance automation and efficiency by enabling seamless data exchange. IoT ecosystem data, generated through intelligent networks like 5G, is growing exponentially. However, its data island nature hinders secure aggregation. Furthermore, cloud-trained models struggle to provide personalized security solutions, which amplifies security and privacy challenges due to increased vulnerability to cyber threats. To address these challenges, deep Federated Learning (FL) solves data island issues without sharing private data, while Cyber Threat Intelligence (CTI) employs AI models to understand and detect cyber threats to protect 5G-enabled IoT ecosystems, which addresses personalization concerns. However, much literature indicates that the privacy of classical and quantum systems can still be compromised during the communication process in deep FL. In addition, the training cost of the local CTI model is also relatively high in deep FL. This study first developed a novel lightweight deep learning-based CTI model that integrates MobileNet-V1 with a bidirectional gated recurrent unit to detect cyber threats. Second, an FL framework is designed to enable multiple IoT clients with heterogeneous data to collectively build a comprehensive CTI model, while preserving privacy and personalized FL. Third, a post-quantum cryptographic protocol based on Learning with Errors is crafted to secure model parameters during client-server communication. Extensive experiments on real-world 5G-NIDD datasets demonstrate the exceptional performance of the proposed model in detecting multiple cyber threats targeting IoT ecosystems, achieving 99.82% accuracy and surpassing contemporary methods. This framework substantially improves protection for the vital 5G-enabled IoT ecosystem against emerging cyber threats in the quantum era.

The increasing adoption of Industrial Internet of Things (IIoT) networks has introduced new security challenges, particularly to ensure software security against evolving malware threats. IIoT systems rely on interconnected edge, cloud, and embedded devices, which are highly vulnerable to malware attacks that exploit software vulnerabilities, propagate across networks, and compromise industrial operations. However, existing malware detection approaches often struggle with resource constraints, high-dimensional feature spaces, and the need for real-time adaptability, making them inefficient for large-scale IIoT deployments. To address these challenges, this paper presents "GWPSO-GAMD," a resource-efficient malware detection framework designed to enhance software security in IIoT networks. The framework integrates a hybrid metaheuristic feature selection algorithm—Grey Wolf Optimization and Particle Swarm Optimization (GWPSO)—to identify the most discriminative and computationally efficient features, reducing processing overhead while maintaining high detection accuracy. These features are then analyzed by the Graph Android Malware Detector (GAMD), which leverages graph convolutional networks (GCNs) and attention mechanisms to model malware propagation behaviors and uncover complex attack patterns in IIoT environments. Empirical evaluations on two open-source malware datasets, CIC-MalDroid-2020 and CIC-MalMem-2022, demonstrate that the proposed model achieves detection accuracy above 98.41% while significantly reducing CPU usage by 47%, memory footprint by 57%, and training time by 67%. The results highlight GWPSO-GAMD’s ability to provide scalable, real-time malware detection for resource-constrained IIoT systems, advancing the vision of secure, intelligent, and resource-aware IIoT networks.

The integration of the Internet of Things (IoT) with traditional healthcare systems has improved quality of healthcare services. However, the wearable devices and sensors used in Healthcare System (HS) continuously monitor and transmit data to the nearby devices or servers using an unsecured open channel. This connectivity between IoT devices and servers improves operational efficiency, but it also gives a lot of room for attackers to launch various cyber-attacks that can put patients under critical surveillance in jeopardy. In this article, a Blockchain-orchestrated Deep learning approach for Secure Data Transmission in IoT-enabled healthcare system hereafter referred to as “BDSDT” is designed. Specifically, first a novel scalable blockchain architecture is proposed to ensure data integrity and secure data transmission by leveraging Zero Knowledge Proof (ZKP) mechanism. Then, BDSDT integrates with the off-chain storage InterPlanetary File System (IPFS) to address difficulties with data storage costs and with an Ethereum smart contract to address data security issues. The authenticated data is further used to design a deep learning architecture to detect intrusion in HS network. The latter combines Deep Sparse AutoEncoder (DSAE) with Bidirectional Long Short-Term Memory (BiLSTM) to design an effective intrusion detection system. Experiments on two public data sources (CICIDS-2017 and ToN-IoT) reveal that the proposed BDSDT outperformed state-of-the-arts in both non-blockchain and blockchain settings and have obtained accuracy close to 99% using both datasets.

The recent burgeoning of Internet of Things (IoT) technologies in the maritime industry is successfully digitalizing Maritime Transportation Systems (MTS). In IoT-enabled MTS, the smart maritime objects, infrastructure associated with ship or port communicate wirelessly using an open channel Internet. The intercommunication and incorporation of heterogeneous technologies in IoT-enabled MTS brings opportunities not only for the industries that embrace it, but also for cyber-criminals. Cyber Threat Intelligence (CTI) is an effective security strategy that uses artificial intelligence models to understand cyber-attacks and can protect data of IoT-enabled MTS proficiently. Unsurprisingly, most of the existing CTI-based solutions uses manual analysis to extract relevant threat information, and has low detection and high false alarm rate. Therefore, to tackle aforementioned challenges, an automated framework called DLTIF is developed for modeling cyber threat intelligence and identifying threat types. The proposed DLTIF is based on three schemes: a deep feature extractor (DFE), CTI-driven detection (CTIDD) and CTI-attack type identification (CTIATI). The DFE scheme automatically extracts the hidden patterns of IoT-enabled MTS network and its output is used by CTIDD scheme for threat detection. The CTIATI scheme is designed to identify the exact threat types and to assist security analysts in giving early warning and adopt defensive strategies. The proposed framework has obtained upto 99% accuracy, and outperforms some traditional and recent state-of-the-art approaches.

The rapid development of the automotive Industrial Internet of Things requires secure networking infrastructure toward digitalization. Cybertwin (CT) is a next-generation networking architecture that serves as a communication, and digital asset owner, and can make the Vehicle-to-Everything (V2X) network flexible and secure. However, CT itself can publish end users’ digital assets to other entities as a service, making data security and privacy major obstacles in the realization of V2X applications. Motivated from the aforementioned discussion, this article presents BDTwin, a blockchain and deep-learning-based integrated framework to enhance security and privacy in CT-driven V2X applications. Specifically, a blockchain scheme is designed to ensure secure communication among vehicles, roadside units, CT-edge server, and cloud server using a smart contract-based enhance-Proof-of-Work (ePoW) and Zero Knowledge Proof (ZKP)-based verification process. Smart contracts are used to enforce rules and regulations that govern the behavior of V2X entities in a nondeniable and automated manner. In a deep-learning scheme, an autoregressive-deep variational autoencoder model is combined with attention-based bidirectional long short-term memory (A-BLSTM) for automatic feature extraction and attack detection by analyzing CT-edge servers data in a V2X environment. Security analysis and experimental results using two different sources, ToN-IoT and CICIDS-2017 show the superiority of the proposed BDTwin framework over some baseline and recent state-of-the-art techniques.

The industrial Internet of Things (IIoT) is a fast-growing network of Internet-connected sensing and actuating devices aimed to enhance manufacturing and industrial operations. This interconnection generates a high volume of data over the IIoT network and raises serious security (e.g., the rapid evolution of hacking techniques), privacy (e.g., adversaries performing data poisoning and inference attacks), and scalability issues. To mitigate the aforementioned challenges, this article presents, a new privacy-preserved threat intelligence framework (P2TIF) to protect confidential information and to identify cyber-threats in IIoT environments. There are two major elements in the proposed P2TIF framework. First, a scalable blockchain module that enables secure communication of IIoT data and prevents data poisoning attacks. Second, a deep learning module that transforms actual data into a new format and protects data from inference attacks using a deep variational autoencoder (DVAE) technique. The encoded data are then employed by a threat detection system using attention-based deep gated recurrent neural network (A-DGRNN) to recognize malicious patterns in IIoT environments. The proposed framework is validated using two different network data sources, i.e., ToN-IoT and IoT-Botnet. Security analysis and experimental results revealed the high efficiency and scalability of the proposed P2TIF framework.

The industrial healthcaresystem has enabled the possibility of realizing advanced real-time monitoring of patients and enriched the quality of medical services through data sharing among intelligent wearable devices and sensors. However, this connectivity brings the intrinsic vulnerabilities related to security and privacy due to the need of continuous communication and monitoring over public network (insecure channel). Motivated from the aforementioned discussions, we integrate permissioned blockchain and smart contract with deep learning (DL) techniques to design a novel secure and efficient data sharing framework named PBDL. Specifically, PBDL first has a blockchain scheme to register, verify (using zero-knowledge proof), and validate the communicating entities using the smart contract-based consensus mechanism. Second, the authenticated data are used to propose a novel DL scheme that combines stacked sparse variational autoencoder (SSVAE) with self-attention-based bidirectional long short term memory (SA-BiLSTM). In this scheme, SSVAE encodes or transforms the healthcare data into new format, and SA-BiLSTM identifies and improves the attack detection process. The security analysis and experimental results using IoT-Botnet and ToN-IoT datasets confirm the superiority of the PBDL framework over existing state-of-the-art techniques.

With the development of Internet of Vehicles (IoV), the integration of Internet of Things (IoT) and manual vehicles becomes inevitable in Intelligent Transportation Systems (ITS). In ITS, the IoVs communicate wirelessly with other IoVs, Road Side Unit (RSU) and Cloud Server using an open channel Internet. The openness of above participating entities and their communication technologies brings challenges such as security vulnerabilities, data privacy, transparency, verifiability, scalability, and data integrity among participating entities. To address these challenges, we present a Privacy-Preserving based Secured Framework for Internet of Vehicles (P2SF-IoV). P2SF-IoV integrates blockchain and deep learning technique to overcome aforementioned challenges, and works on two modules. First, a blockchain module is developed to securely transmit the data between IoV-RSU-Cloud. Second, a deep learning module is designed that uses the data from blockchain module to detect intrusion and its performance is assessed using two network datasets IoT-Botnet and ToN-IoT. In contrast with other peer privacy-preserving intrusion detection strategies, the P2SF-IoV approach is compared, and the experimental results reveal that in both blockchain and non-blockchain based solutions, the proposed P2SF-IoV framework outperforms.

Cooperative Intelligent Transport System (C-ITS) is a promising technology that aims to improve the traditional transport management systems. In C-ITS infrastructure Autonomous Vehicles (AVs) communicate wirelessly with other AVs, Road Side Units (RSUs) and Traffic Command Centres (TCCs) using an open channel Internet. However, the use of the Internet brings inherent vulnerabilities related to privacy (e.g., adversary performing inference and data poisoning attacks), and security (e.g., AVs can be compromised using advanced hacking techniques) issues and prevents the faster realization of C-ITS applications. To address these challenges, this paper presents a privacy-preserving-based secure framework to provide both privacy and security in C-ITS infrastructure. The proposed framework provides two level of security and privacy using blockchain and deep learning modules. First, a blockchain module is designed to securely transmit the C-ITS data between AVs–RSUs-TCCs, and a smart contract-based enhanced Proof of Work (ePoW) technique is designed to verify data integrity and mitigate data poisoning attacks. Second, a deep-learning module is designed that includes Long-Short Term Memory-AutoEncoder (LSTM-AE) technique for encoding C-ITS data into a new format to prevent inference attacks. The encoded data is used by the proposed Attention-based Recurrent Neural Network (A-RNN), for intrusive events recognition in C-ITS infrastructure. The proposed A-RNN is trained using Truncated Backpropagation Through Time (BPTT) algorithm. The framework is further validated and tested using two publicly available ToN-IoT and CICIDS-2017 datasets. The proposed framework is compared with peer privacy-preserving intrusion detection techniques, and the result shows the effectiveness of the proposed framework over several state-of-the-art techniques in both blockchain and non-blockchain systems.

Localization problem has gained a significant attention in the field of wireless sensor networks in order to support location-based services or information such as supporting geographic routing protocols, tracking events, targets, and providing security protection techniques. A number of variants of DV-Hop-based localization algorithms have been proposed and their performance is measured in terms of localization error. In all these algorithms, while determining the location of a non-anchor node, all the anchor nodes are taken into consideration. However, if only the anchors close to the node are considered, it will be possible to reduce the localization error significantly. This paper explores the effect of the close anchors in the performance of the DV-Hop-based localization algorithms and an improvement is proposed by considering only the closest anchors. The simulation results show that considering closest anchors for estimation of the location reduces localization error significantly as compared to considering all the anchors.