A Convolutional Neural Network for Electrical Fault Recognition in Active Magnetic Bearing Systems.

Active magnetic bearings are complex mechatronic systems that consist of mechanical, electrical, and software parts, unlike classical rolling bearings. Given the complexity of this type of system, fault detection is a critical process. This paper presents a new and easy way to detect faults based on the use of a fault dictionary and machine learning. The dictionary was built starting from fault signatures consisting of images obtained from the signals available in the system. Subsequently, a convolutional neural network was trained to recognize such fault signature images. The objective of this study was to develop a fault dictionary and a classifier to recognize the most frequent soft electrical faults that affect position sensors and actuators. The proposed method permits, in a computationally convenient way that can be implemented in real time, the determination of which component has failed and what kind of failure has occurred. Therefore, this fault identification system allows determining which countermeasure to adopt in order to enhance the reliability of the system. The performance of this method was assessed by means of a case study concerning a real turbomachine supported by two active magnetic bearings for the oil and gas field. Seventeen fault classes were considered, and the neural network fault classifier reached an accuracy of 93% on the test dataset.

Improving CSI Prediction Accuracy with Deep Echo State Networks in 5G Networks.

The forthcoming fifth-generation networks require improvements in cognitive radio intelligence, going towards more smart and aware radio systems. In the emerging radio intelligence approach, the empowerment of cognitive capabilities is performed through the adoption of machine learning techniques. This paper investigates the combined application of the convolutional and recurrent neural networks for the channel state information forecasting, providing a multivariate scalar time series prediction by taking into account the multiple factors dependence of the channel state conditions. Finally, the system performance has been analyzed in terms of prediction accuracy expressed as absolute deviation error and mean percentage error, in comparison with an alternative machine learning method recently proposed in the literature with the aim at solving the same prediction problem.

Toward Resilient Wireless Sensor Networks: A Virtualized Perspective.

The Internet of Things (IoT) has been one of the main focus areas of the research community in recent years, the requirements of which help network administrators to design and ensure the functionalities and resources of each device. Generally, two types of devices-constrained and unconstrained devices-are typical in the IoT environment. Devices with limited resources-for example, sensors and actuators-are known as constrained devices. Unconstrained devices includes gateways or border routers. Such devices are challenging in terms of their deployment because of their connectivity, channel selection, multiple interfaces, local and global address assignment, address resolution, remote access, mobility, routing, border router scope and security. To deal with these services, the availability of the IoT system ensures that the desired network services are available even in the presence of denial-of-service attacks, and the use of the system has become a difficult but mandatory task for network designers. To this end, we present a novel design for wireless sensor networks (WSNs) to address these challenges by shifting mandatory functionalities from unreliable to reliable and stable domains. The main contribution of our work consists in addressing the core network requirements for IoT systems and pointing out several guidelines for the design of standard virtualized protocols and functions. In addition, we propose a novel architecture which improves IoT systems, lending them more resilience and robustness, together with highlighting and some important open research topics.

False Data Detection for Fog and Internet of Things Networks.

The Internet of Things (IoT) context brings new security issues due to billions of smart end-devices both interconnected in wireless networks and connected to the Internet by using different technologies. In this paper, we propose an attack-detection method, named Data Intrusion Detection System (DataIDS), based on real-time data analysis. As end devices are mainly resource constrained, Fog Computing (FC) is introduced to implement the DataIDS. FC increases storage, computation capabilities, and processing capabilities, allowing it to detect promptly an attack with respect to security solutions on the Cloud. This paper also considers an attack tree to model threats and vulnerabilities of Fog/IoT scenarios with heterogeneous devices and suggests countermeasure costs. We verify the performance of the proposed DataIDS, implementing a testbed with several devices that measure different physical quantities and by using standard data-gathering protocols.