The proliferation of smart devices in the industrial Internet of Things (IIoT) will generate large volumes of data traffic in the 5G era. Encrypting network traffic helps to secure the data security of smart devices while also raising the complexity of classification. Due to the issue of traditional machine learning methods depending on expert experience and the fact that most current research only uses a single CNN and RNN network to detect encrypted traffic without taking forward and backward correlations between data packets into consideration, it is challenging to identify encrypted traffic features effectively. This paper proposes CNN-BiLSTM, a method that integrates spatiotemporal characteristics with attention mechanisms. In particular, we utilize 1D-CNN and BiLSTM to extract spatial and temporal features from encrypted traffic; we then apply the CBAM block and TPA block to weight the extracted features, allowing the model to concentrate on highly discriminative features; additionally, we applied the ELU activation function in the CNN model to enhance the model's convergence speed and avoid gradient vanishing issues. This study was validated experimentally using the ISCX2016 public dataset, and the experimental findings demonstrate that, in comparison to current mainstream methods, the model performs exceptionally well in categorization.

Three prominent traffic features including peak alignment, stable ranking, and gravity model, have guided the design of current Google Jupiter fabrics in traffic engineering, topology engineering, and capacity planning.

The expansion of the scale of the Power Internet of Things stimulated by the development of the Energy Internet makes the growth in demand for the effective authentication and access control technologies in the cross-domain data exchange. Based on the cross-chain technology of the blockchain and the cuckoo filter, this paper proposes a cross-domain authentication scheme for Power Internet of Things. Firstly, a cross-chain authentication architecture is established. Combined with the existing authentication technologies used in intra-domain authentication, a cross-domain authentication process based on the cross-chain technology is proposed to realize the automatic transmission of the authentication credentials from application domain to authentication domain. The cuckoo filter is deployed on the blockchain as smart contracts, and the user certificate fingerprint is inserted into the filter to realize user registration, query, and revocation, which reduces the cost of the user certificate management. Experimental results show the effectiveness and feasibility of our scheme. Based on the proposed authentication scheme, a cross-domain access control scheme based on roles and object classes is presented, by treating the object classes as controlled objects and then applying the role-based access control to the object classes, on the condition that the heterogeneous domains in the Energy Internet have the same kinds of resources.

In recent years, there has been a noticeable surge in electric power load due to economic development and improved living standards. The growing need for smart power solutions, such as leveraging user electricity data to forecast power peaks and utilizing power data statistics to enhance end-user services, has been on the rise. However, the misuse and unauthorized access of data have prompted stringent regulations to safeguard data integrity. This paper presents a novel decentralized collaborative machine learning framework aimed at predicting peak power loads while protecting the privacy of users’ power data. In this scheme, multiple users engage in collaborative machine learning training within a peer-to-peer network free from a centralized server, with the objective of predicting peak power loads without compromising users’ local data privacy. The proposed approach leverages blockchain technology and advanced cryptographic techniques, including multi-key homomorphic encryption and consistent hashing. Key contributions of this framework include the development of a secure dual-aggregate node aggregation algorithm and the establishment of a verifiable process within a decentralized architecture. Experimental validation has been conducted to assess the feasibility and effectiveness of the proposed scheme, demonstrating its potential to address the challenges associated with predicting peak power loads securely and preserving user data privacy.

Machine learning algorithms are becoming increasingly common in malware detection. However, these algorithms typically focus only on optimizing model accuracy and recall rate, without fully considering the fact that attackers may intentionally make changes to the malware so that generated adversarial malware can evade malware detectors while preserve the malicious functionality. This paper proposed a black-box adversarial malware generation method with Reinforcement Learning based on proximal policy optimization. The method takes actions based on learned policy to make changes to the origin malware. Compared to previous methods, our approach employed the PPO algorithm to enhance the stability of the model training process, and a more reasonable action space is designed to modify the origin malware. The experiment results show that it has over 94.8%-97.4% evasion rate for three state-of-the-art machine learning based detectors in a pure black-box setting.