Fang Fan

With the rapid advancement of artificial intelligence, the Internet of Things (IoT), and related technologies, cybersecurity concerns are escalating. Reconstruction attacks pose a substantial threat to the privacy and security of machine learning models. This paper explores innovative approaches, specifically federated learning (FL) and generative adversarial networks (GAN), to bolster intrusion detection capabilities. Federated learning, a decentralized machine learning paradigm, empowers individual network clients to conduct model training locally, mitigating privacy risks associated with centralized data uploads. While FL addresses data privacy concerns, it falls short in protecting against reconstruction attacks—malicious attempts to reconstruct sensitive data from training models.To tackle this vulnerability, we integrate GAN into the FL framework, presenting a novel method for IoT anomalous traffic intrusion detection. By combining FL's decentralized training with GAN's generative capabilities, our approach achieves heightened accuracy in identifying and thwarting reconstruction attacks. Experimental evaluations, utilizing publicly available datasets, showcase the superior performance of our method compared to traditional intrusion detection techniques.Our method not only enhances data privacy but also introduces a paradigm shift in network security research and applications. This paper contributes valuable insights and methods, opening new avenues for advancing the field of intrusion detection in the era of evolving cyber threats.

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.

The Industrial Internet of Things (IIoT) represents the deployment of Internet of Things (IoT) technology in industrial applications. In this article, we address the challenges of fault diagnosis and data privacy protection within the IIoT environment. We present a novel fault diagnosis model that combines federated learning and a particle swarm optimization algorithm. Firstly, we introduce a three-tier federated learning model designed to safeguard the data privacy of each participant in a real industrial control network structure. Subsequently, we enhance the particle swarm optimization algorithm to augment its global exploration capabilities and convergence performance, enabling it to collect federated learning model weights in lieu of traditional techniques. Furthermore, we employ the Taguchi method to tailor an optimal solution for the modified particle swarm optimization algorithm (TMPSO), thereby enhancing the algorithm's efficiency and robustness. Additionally, we propose a neural network model utilizing small convolutional kernels (SVGG) for fault diagnosis within the IIoT framework, thereby improving the model's feature learning capabilities. Experimental validation was conducted using actual industrial rolling bearing datasets and CIFAR-10 datasets. The results of these experiments demonstrate that our proposed TPMPSO-SVGG model outperforms other methods in terms of fault identification accuracy and communication cost.

A quasi-affine transformation evolutionary algorithm improved by the Taguchi strategy, levy flight and the restart mechanism (TLR-QUATRE) is proposed in this paper. This algorithm chooses the specific optimization route according to a certain probability, and the Taguchi strategy helps the algorithm achieve more detailed local exploitation. The latter two strategies help particles move at random steps of different sizes, enhancing the global exploration ability. To explore the new algorithm’s performance, we make a detailed analysis in seven aspects through comparative experiments on CEC2017 suite. The experimental results show that the new algorithm has strong optimization ability, outstanding high-dimensional exploration ability and excellent convergence. In addition, this paper pays attention to the demonstration of the process, which makes the experimental results credible, reliable and explainable. The new algorithm is applied to fault detection in wireless sensor networks, in which TLR-QUATRE is combined with back-propagation neural network (BPNN). This study uses the symmetry of generation and feedback for network training. We compare it with other optimization structures through eight public datasets and one actual landing dataset. Five classical machine learning indicators and ROC curves are used for visualization. Finally, the robust adaptability of TLR-QUATRE on this issue is confirmed.

Affected by solar radiation, atmospheric windows, radiation aberrations, and other air and sky environmental factors, remote sensing images usually contain a large amount of noise and suffer from problems such as non-uniform image feature density. These problems bring great difficulties to the segmentation of high-precision remote sensing image. To improve the segmentation effect of remote sensing images, this study adopted an improved metaheuristic algorithm to optimize the parameter settings of pulse-coupled neural networks (PCNNs). Using the Taguchi method, the optimal parallelism scheme of the algorithm was effectively tailored for a specific target problem. The blindness in the design of the algorithm parallel structure was effectively avoided. The superiority of the customized parallel SCA based on the Taguchi method (TPSCA) was demonstrated in tests with different types of benchmark functions. In this study, simulations were performed using IKONOS, GeoEye-1, and WorldView-2 satellite remote sensing images. The results showed that the accuracy of the proposed remote sensing image segmentation model was significantly improved.

Aiming at the intrusion detection problem of the wireless sensor network (WSN), considering the combined characteristics of the wireless sensor network, we consider setting up a corresponding intrusion detection system on the edge side through edge computing. An intrusion detection system (IDS), as a proactive network security protection technology, provides an effective defense system for the WSN. In this paper, we propose a WSN intelligent intrusion detection model, through the introduction of the k-Nearest Neighbor algorithm (kNN) in machine learning and the introduction of the arithmetic optimization algorithm (AOA) in evolutionary calculation, to form an edge intelligence framework that specifically performs the intrusion detection when the WSN encounters a DoS attack. In order to enhance the accuracy of the model, we use a parallel strategy to enhance the communication between the populations and use the Lévy flight strategy to adjust the optimization. The proposed PL-AOA algorithm performs well in the benchmark function test and effectively guarantees the improvement of the kNN classifier. We use Matlab2018b to conduct simulation experiments based on the WSN-DS data set and our model achieves 99% ACC, with a nearly 10% improvement compared with the original kNN when performing DoS intrusion detection. The experimental results show that the proposed intrusion detection model has good effects and practical application significance.

Aiming at the problem of fault detection in data collection in wireless sensor networks, this paper combines evolutionary computing and machine learning to propose a productive technical solution. We choose the classical particle swarm optimization (PSO) and improve it, including the introduction of a biological population model to control the population size, and the addition of a parallel mechanism for further tuning. The proposed RS-PPSO algorithm was successfully used to optimize the initial weights and biases of back propagation neural network (BPNN), shortening the training time and raising the prediction accuracy. Wireless sensor networks (WSN) has become the key supporting platform of Internet of Things (IoT). The correctness of the data collected by the sensor nodes has a great influence on the reliability, real-time performance and energy saving of the entire network. The optimized machine learning technology scheme given in this paper can effectively identify the fault data, so as to ensure the effective operation of WSN.

The wide application of wireless sensor networks (WSN) brings challenges to the maintenance of their security, integrity, and confidentiality. As an important active defense technology, intrusion detection plays an effective defense line for WSN. In view of the uniqueness of WSN, it is necessary to balance the tradeoff between reliable data transmission and limited sensor energy, as well as the conflict between the detection effect and the lack of network resources. This paper proposes a lightweight Intelligent Intrusion Detection Model for WSN. Combining k-nearest neighbor algorithm (kNN) and sine cosine algorithm (SCA) can significantly improve the classification accuracy and greatly reduce the false alarm rate, thereby intelligently detecting a variety of attacks including unknown attacks. In order to control the complexity of the model, the compact mechanism is applied to SCA (CSCA) to save the calculation time and space, and the polymorphic mutation (PM) strategy is used to compensate for the loss of optimization accuracy. The proposed PM-CSCA algorithm performs well in the benchmark functions test. In the simulation test based on NSL-KDD and UNSW-NB15 data sets, the designed intrusion detection algorithm achieved satisfactory results. In addition, the model can be deployed in an architecture based on cloud computing and fog computing to further improve the real-time, energy-saving, and efficiency of intrusion detection.

The key to the problem of node coverage in wireless sensor networks (WSN) is to deploy a limited number of sensors to achieve maximum coverage. This paper studies the hybrid strategies of multiple evolutionary algorithms, and applies them to the problem of WSN node coverage. We first proposed the hybrid algorithm SFLA‐WOA (SWOA) based on Shuffled Frog Leaping Algorithm (SFLA) and Whale Optimization Algorithm (WOA). The SWOA algorithm combines the advantages of SFLA and WOA; that is, it retains the unique evolution model of WOA and also has the excellent co‐evolution capability of SFLA. Secondly, using the mutation, crossover and selection operations of the differential evolution (DE) algorithm to further optimize this hybrid algorithm, the SWOA‐based SFLA‐WOA‐DE (SWOAD) algorithm is proposed. In addition, the performance of SWOA and SWOAD has been tested by 30 benchmark functions in the CEC 2017 test set. Experimental results show that the optimization effects of these two algorithms are very outstanding. Finally, the simulation results show that the optimization algorithm proposed in this paper has a good effect on improving the signal coverage of WSN under the actual three‐dimensional terrain.