With the widespread use of Internet of things(IoT) devices, it generates an enormous volume of data, and it is a challenge to mine the IoT data value while ensuring security and privacy. Federated learning is a decentralized approach for training data located on edge devices, such as mobile phones and IoT devices, while keeping privacy, efficiency, and security. However, the Non-IID (non-independent and identically distributed) data, always greatly impacts the performance of the global model. In this paper, we propose a FedDynamic algorithm to solve the statistical challenge of federated learning caused by Non-IID. As Non-IID data can lead to significant differences in model parameters between edge devices, we set different weights for different devices during model aggregation to get a high-performance global model. We analyze and exact key indices (local model accuracy, local data quality, and model difference between local models and the global model), which can reflect the quality of the model, and calculate the aggregation weight for edge devices based on the key indices. Furthermore, we dynamically adjust aggregation weight based on accuracy’s variety to solve weight staleness during the training process. Experiments on the MNIST, FMNIST, EMNIST, CINIC-10 and CIFAR-10 datasets show that the FedDynamic algorithm has better accuracy and convergence performance, compared to the FedAvg, FedProx and Scaffold algorithms.

With the continuous development of technologies such as the Internet of Things (IoT) and cloud computing, sensors collect and store large amounts of sensory data, realizing real-time recording and perception of the environment. Due to the open characteristics of WSN, the security risks during information transmission are prominent, and network attack or intrusion is likely to occur. Therefore, effective anomaly detection is vital for IoT systems to keep the system safe. The original Isolation Forest algorithm is an anomaly detection algorithm with linear time complexity and has a better detection effect on perceptual data. However, there are also disadvantages such as strong randomness, low generalization performance, and insufficient stability. This paper proposes a data anomaly detection method named BS-iForest (box plot-sampled iForest) for wireless sensor networks based on a variant of Isolation Forest to address the problems. This method first uses the sub-dataset filtered by the box graph to train and construct trees. Then, isolation trees with higher accuracy are selected in the training set to form a base forest anomaly detector. Next, the base forest anomaly detector uses anomaly detection to judge data outliers during the next period. These experiments were performed on datasets collected from sensors deployed in a data center of a university, and the Breast Wisconsin (BreastW) dataset, showing the performance of the variant of the Isolation Forest algorithm. Compared with the traditional isolation forest, the area under the curve (AUC) increased by 1.5% and 7.7%, which verified that the proposed method outperforms the standard Isolation Forest algorithm with the two datasets we chose.

People connected by huge-size socialnetworks are highly dependent on recommendation systems to discover interesting persons, contents, and commodities. Extracting user interests and modeling their evolutions from user historical behavior are vital for algorithms to judge whether users are interested in given persons and items. The existing methods have two shortcomings: firstly, they do not have the high-order ability to process temporal sequences, which leads to their inability to mine the evolution of user interests; secondly, the models take full large-scale information of items as input, which causes a severe problem of overfitting. Thus, we propose a new hybrid neural network, DGRU, which integrates Factorization-Machine Based Neural Network (DeepFM) and Gated Recurrent Unit Neural Network (GRU). The DeepFM component is responsible for performing the autonomic feature combination, and the GRU component is designed to model user interests and evolutions. The GRU component is fed with a series of 1 and 0 representing user click behaviors. It contains information on what users like and dislike. Moreover, the conciseness of the format is helpful to avoid the problem of overfitting. Experiments on three real datasets demonstrated that the proposed model has a better CTR prediction performance and robustness than the state-of-the-art models.

With the increase in intelligence applications and services, like real-time video surveillance systems, mobile edge computing, and Internet of things (IoT), technology is greatly involved in our daily life. However, the reliability of these systems cannot be always guaranteed due to the hard disk drive (HDD) failures of edge nodes. Specifically, a lot of read/write operations and hazard edge environments make the maintenance work even harder. HDD failure prediction is one of the scalable and low-overhead proactive fault tolerant approaches to improve device reliability. In this paper, we propose an LSTM recurrent neural network-based HDD failure prediction model, which leverages the long temporal dependence feature of the drive health data to improve prediction efficiency. In addition, we design a new health degree evaluation method, which stores current health details and deterioration. The comprehensive experiments on two real-world hard drive datasets demonstrate that the proposed approach achieves a good prediction accuracy with low overhead.

An intelligent maritime navigation system is expected to play an important role in the realm of Internet of Vessels (IoV). As a key technology in navigation systems, vessel trajectory prediction technology is critical to the IoV. Automatic identification system (AIS), an automated tracking system, is used extensively for vessel trajectory prediction. However, certain characteristics in the AIS data, such as the large number of anchored trajectories in the area, anomalous sharp turns of some trajectories, and the behavioral differences of vessels in different segments, limit the prediction accuracy. In this study, we propose a novel vessel trajectory prediction model for accurate prediction with the following characteristics: (1) an anchor trajectory elimination algorithm to eliminate anchor trajectories; (2) a statistical trajectory restoration algorithm to repair sharp turning; (3) a two-stage clustering algorithm (D-KMEANS) to distinguish vessel behavior; and (4) a deep bidirectional gate recurrent unit (Stacked-BiGRUs) model to predict vessel trajectory and compare the accuracy of the model before and after improvement. The results show that the mean square error and the mean absolute error of the improved model are reduced by 27% and 46%, respectively. This research shows good potential for maritime navigation early warning and safety.