Abstract In recent years, deep learning-based emotion recognition from electroencephalography (EEG) signals has garnered significant attention in brain-computer interfaces. However, effectively capturing local and global dependencies remains a challenge due to the complexities of EEG data. Furthermore, traditional convolutional neural networks and RNNs often struggle to fully explore the spatio-temporal relationships between different features. To address these issues, we propose an end-to-end model with the augmented capsule-gated Transformer to improve the performance of EEG emotion recognition, in which we learn cross-channel spatial features effectively, and the raw EEG signals are automatically weighted to emphasize key attributes. Subsequently, the capsule network extracts low-level and high-level spatial information, fully leveraging the potential insights within the signals. Building on this, an efficient Transformer is employed to model the relationships among different electrodes, allowing for a more in-depth analysis of the temporal dependencies across multiple features. Extensive experiments are conducted on the Dataset for Emotion Analysis using Physiological Signals (DEAP) dataset, and comparison results with existing state-of-the-art methods demonstrate the superior performance of the proposed method. Specifically, for the arousal and valence dimensions, the average recognition accuracies in subject-dependent experiments reach 93.51% and 94.24%, while the subject-independent experiments achieve average accuracies of 86.78% and 87.59%.

With the proliferation of Internet of Things (IoT) devices, vast amounts of sensitive data are frequently exchanged across networks, making identity authentication crucial to secure communication. However, existing authentication schemes generally suffer from complex certificate management, difficult key custody, vulnerability to various attacks, and high overhead, making them unsuitable for resource-constrained IoT devices. Despite blockchain technology showing promise in decentralized authentication, consensus mechanisms’ high computational and latency costs hinder their application in cross-domain environments. To address these challenges, this paper proposes a lightweight cross-domain authentication and key agreement protocol, termed LCH-AKA, designed for heterogeneous IoT systems. LCH-AKA integrates Physical Unclonable Function (PUF), blockchain, and edge computing technologies to construct a decentralized and tamper-resistant authentication framework. The proposed scheme eliminates the need for traditional certificates and centralized key management, while enabling efficient device-to-device and device-to-server mutual authentication. Formal verification and experimental evaluation demonstrate that LCH-AKA achieves strong security, scalability, and low resource consumption. Compared with existing approaches, it provides lower computational and communication overhead, reduced latency, and improved energy efficiency, making it suitable for large-scale deployment in heterogeneous IoT networks.

Trajectory prediction is crucial for ensuring the safety and reliability of autonomous driving systems. Accurately predicting the future trajectory of a vehicle can aid drivers or automated driving systems in developing reasonable movement plans and avoiding potential collisions. In real-world scenarios, drivers selectively observe areas that influence their driving decisions based on real-time traffic scenarios, adapting to changing traffic situations to ensure safety. To simulate this behavior, we propose a selective perception module that utilizes historical trajectories and lane centerlines to extract vital information from the changing traffic scene. Furthermore, vehicles tend to follow lane guidance to some extent, we use a normalized flow decoder and small-scale guided loss to induce the distribution of predicted trajectories, slightly increasing the probability of lane centerlines in the predictions as a weak lane guidance. We conduct ablation studies using the public dataset nuScenes and comparative experiments with state-of-the-art methods, showcasing the predictive diversity of our approach while maintaining high accuracy.

The deep integration of Internet of Medical Things (IoMT) and Artificial intelligence makes the further development of intelligent medical services possible, but privacy leakage and data security problems hinder its wide application. Although the combination of IoMT and federated learning (FL) can achieve no direct access to the original data of participants, FL still can't resist inference attacks against model parameters and the single point of failure of the central server. In addition, malicious clients can disguise as benign participants to launch poisoning attacks, which seriously compromises the accuracy of the global model. In this paper, we design a new privacy protection framework (BFG) for decentralized FL using blockchain, differential privacy and Generative Adversarial Network. The framework can effectively avoid a single point of failure and resist inference attacks. In particular, it can limit the success rate of poisoning attacks to less than 26%. Moreover, the framework alleviates the storage pressure of the blockchain, achieves a balance between privacy budget and global model accuracy, and can effectively resist the negative impact of node withdrawal. Simulation experiments on image datasets show that the BFG framework has a better combined performance in terms of accuracy, robustness and privacy preservation.

With the proposal of the intelligent transportation system, vehicular ad-hoc networks have been widely concerned and well-developed. Vehicular ad-hoc network is recognized as a major innovation of the Internet of Things technology. It can collect and analyze traffic data uploaded by vehicles through the network, monitor vehicle state in real-time, provide better driving routes and real-time traffic decisions for vehicles, and improve vehicles’ overall level of intelligent driving. To make vehicles quickly join the vehicular ad-hoc networks through the authentication of identity legitimacy, take into account security and efficiency, and further reduce the calculation and communication consumption in authentication, this paper designs a mutual anonymous authentication and key agreement scheme based on an elliptic curve for the vehicular ad-hoc networks. To complete vehicle authentication and session key establishment, we work with lightweight operations like hash, XOR, and connection in conjunction with the elliptic curve discrete logarithm problem to guarantee the confidentiality of communication data. In the scheme, identity authentication is divided into two types: initial authentication and subsequent authentication. When the vehicle is just on the road, it will use the first roadside unit it encounters for initial authentication. All roadside units that cars on the road come across are then authenticated after the accomplishment of the initial authentication with the first roadside unit. Subsequent authentication is lighter and less computationally complex than initial authentication. Of course, subsequent authentication is based on initial authentication. This paper also analyzes the scheme’s security and uses BAN logic analysis and Proverif simulation to verify the scheme’s security. Additionally, performance analysis is used to demonstrate the scheme’s superiority.

Clustering, as one of the main tasks of machine learning, is also the core work of granular computing, namely granulation. Most of the recent granular computing based clustering algorithms only utilize the plain granule features without taking the granule structure into account, especially in information area with widespread application of algebraic structure. This paper aims at proposing an algebraic structure based clustering method from granular computing prospective. Specifically, the algebraic structure based granularity is firstly formulated based on the granule structure of an algebraic binary operator. An algebraic structure based clustering method is then proposed by incorporating congruence partitioning granules and homomorphically projecting granule structure. Finally, proof of the lattice at multiple hierarchical levels and comparative analysis of experimental cases validate the effectiveness of the proposed clustering method. The algebraic structure based clustering method can provide a general framework to perform granularity clustering using the algebraic granule structure information. It meanwhile advances the granular computing methods by combing the granular computing theory and the clustering theory.

The selfish On-Board-Unit (OBU) attacks Vehicular Ad-Hoc Network (VANET) by various attacks for profit. However, many existing methods are based on the principle of direct reciprocity for communication, and when an attack occurs, it is easy to crash in the case of large-scale networks. In order to reduce the number of attackers in the vehicle ad-hoc network and restrain the attack motivation of the OBUs, we propose an indirect reciprocal incentive mechanism based on reputation to encourage the OBUs in the VANET to help each other. Since most OBUs are in great need of network services, including potential attackers, when the loss of network services is far greater than the illegal benefits of their attacks, selfish and rational OBU will give up attacks and take desirable behavior. In addition, to prevent some attacks from tampering with information, we also apply blockchain technology to record the behavior of OBU. The indirect reciprocity process of each OBU in VANET can be regarded as a Markov Decision Process (MDP). In order to restrain the attack motivation of selfish nodes and communicate normally without knowing the attack model, an algorithm based on Deep Reinforcement Learning (DRL) is proposed to suppress attack motivation, so as to activate OBU learning in dynamic environment and make wise decisions. Finally, through a large number of simulation experiments, the performance of our proposed algorithm is obviously better than that of the baseline strategy, and is verified by the simulation results.

This Special Collection aims at seeking original articles with novel perspectives and solutions to address the cryptographic security and information hiding technology for Cloud or Fog-based IoT system. We expect this Special Collection can provide scientists, researchers, and industrial practitioners with a chance to publish original manuscripts that demonstrate and explore current advances in all aspects of security, privacy, trust and covert communication issue for Cloud or Fog computing/architecture IoT system.

Against the backdrop of the booming supply chain finance and logistics industry, Logistics 4.0 has emerged. However, the financing capacity of today's logistics companies is still unable to respond to the needs of the rapid development of the supply chain. On the one hand, the problem of logistics companies’ lack of existing high-value collateral in supply chain finance has led to a clutter of their credit data and difficulty in verifying their creditworthiness. On the other hand, the massive access to logistics companies’ business information can also lead to privacy leaks. At the same time, the transparency feature of blockchain is used to solve the financing dilemma of many industries in supply chain finance. On this basis, we propose a blockchain-based financing scheme (BFS) for logistics company. BFS utilises more efficient and interpretative smart contract technology, as well as improved privacy information query and invocation algorithms, to realise automatic control of entity node privacy information flow and significantly simplifying the steps and lowering the thresholds to financing for logistics companies, while protecting the privacy of their data. After extensive simulation testing, BFS can run supply chain blocks at a stable transaction throughput of around 280 RPS, with data transfers that meet the financing needs of logistics companies, providing a higher and more stable performance than the native Hyperledger Fabric.

The Internet of Vehicles is deployed in an open environment, and protecting its security and data privacy is the challenge. Carrying out the Internet of Vehicles secure authentication before interaction of information is an important part of ensuring the security foundation. Therefore, this article designs a safe and reliable Internet of Vehicles authentication and key agreement schemeassisted by blockchain. This article uses a multi-TA network model to improve the efficiency of authentication. Because of the rapid movement of vehicles, it will continue to appear cross-RSU and TA certification. Considering the disadvantages of most centralised authentication protocols using a single TA, this paper uses the multi-TA model to improve the efficiency of authentication. By usingblockchain technology to store the authentication information of vehicles, the cross-domain authentication of vehicles and the protection of user privacy information can be well realised. At the same time, in order to reduce the time of vehicle authentication, this scheme uses a lightweight calculation operation to complete the whole process of authentication. Through security analysis and results of Proverif simulation, the security of the solution is well proved, our scheme can resist various common attacks. Compared with some existing Internet of Vehicles security authentication protocols, the proposed scheme has a lower cost of computation, communication, and storage.

In recent years, with the increase of industrial volume, problems such as difficult management and low efficiency have become obstacles to the development of industry. Many studies have shown that industry can complete the automation of industry by combining with Internet of Things, that is, Industrial Internet of Things (IIoT), which will contribute to the secondary development of industry. However, in IIoT, there are also malicious attacks against industrial devices or sensors, such as DDoS, Sybil, etc., which will cause great security risks to the system. As a popular security trust framework, blockchain technology will face the dilemma of low throughput and limited energy if it is directly applied in the IoT devices. Based on this, in view of the existing IIoT security problems and the difficulties of traditional blockchain based on chain application, this paper proposes a blockchain security solution based on Directed Acyclic Graph(DAG) structure for IIoT, and combines differential privacy technology to further ensure the privacy and integrity of data. In order to ensure the stability and anti-interference ability of the network, this paper also proposes a load balancing algorithm, which effectively balances the relationship between node power consumption and network lifetime. Finally, the simulation results show that the proposed scheme is superior to the chain-blockchain scheme in network performance and power consumption. The average network delay is reduced by 24%, and the average network lifetime is increased by 48%.