Mobile Wireless Middleware, Operating Systems and Applications

In recent years, the aging of the population has become a social problem. Under such circumstances, domestic accidents have become a problem, and these accidents have a severe impact on the health of the elderly. Therefore, a system to watch over the elderly is needed to prevent accidents and detect accidents. However, the current practical system using camera images causes privacy problems for the target of the watch. In this study, we propose a behavior estimation system that uses micro-vibration sensors and deep Boltzmann machines to overcome the issue of privacy and the noise that micro-vibration sensors are prone to. The plan was trained for five behavioral patterns: normal walking, limping, jumping with both feet, jogging, and no action, and then behavior estimation experiments were conducted on the time series measured data for the five behavioral patterns. The experimental results showed that the estimation accuracy was about 72%, confirming the effectiveness of the proposed system. This result suggests that the proposed method can be an essential element for constructing a privacy-conscious watchdog system.

Recently, the basic concept of cyber-physical systems has been extended to various research fields such as cyber-physical-social systems and cyber-physical-human systems. Thanks to the rapid progress of network technology, we can realize real-time tracking in a wide area. We can use features and structures based on graph theory and topology to predict the future state based on real-time tracking. For big data problems, growing neural gas (GNG) can cope well with hidden topological logic features. In this paper, we propose an analysis method of topological tracking based on GNG. First, we apply multi-scale batch learning GNG (MS-BL-GNG) to extract the topological structure for tracking. The proposed method can be used to support the navigation task of the robot, and the proposed method can be used to analyze the topological tracking in motion. We conducted simulation experiments for the proposed method. The experimental results fully demonstrate the effectiveness of the proposed method.

This chapter aims to bring a metric of capacities and applicability of Light Fidelity (Li-Fi) as a communication system through a direct comparison to subareas of Wireless Fidelity (Wi-Fi) communication that, actually, is the main wireless communication system. The actual vision of Li-Fi in a global scenario was discussed based on the use cases of this technology and your relation to 5th-generation (5G) technology. Finally, the difficulties and new paths that this system must still take were presented.

In wireless sensor networks, the radio of the wireless sensor node happens to be the highest source of energy consumption. Hence, there is a need to focus on the MAC layer, as it controls access to the radio. While there are several existing techniques to make sensors more energy-efficient, not many of them consider the security aspects of energy efficiency. By this we mean protecting energy from external attacks. The existing protocols focus mainly on either duty cycling (Sensor-MAC, Time-out MAC) or clustering (Gateway MAC), as a way of conserving energy. One of such attacks to energy is the denial-of-sleep (DoSL) attack which is a specific kind of denial-of-service attacks designed to drain the energy of battery-powered sensors in a wireless sensor network. This paper explains the development of a new MAC layer protocol called Layered-MAC aimed at not just energy efficiency but energy protection against DoSL attacks. The protocol is implemented on the OMNET++ and Castalia simulator. The results from the simulation are then compared with two representative existing duty-cycled protocols (Time-out MAC and Sensor-MAC), and significant improvements are present. One of the benefits of the developed protocol is that not only does it attempt to save energy, but it protects energy from DoSL attacks. There are two main contributions from this research – the first is the additional layer of network metrics (RSSI and LQI) consideration, based on the premise that protection/security is not possible without some form of measurement of assets, and the cluster head rotation which adds an extra layer of energy protection while considering energy efficiency.

The distribution and change of travel intensity reflect the pattern of the city and the activity of trip population. It is important to understand the pattern of the city and the activity of trip flow for urban planning and government decision-making. This paper constructs a Bayesian hierarchical spatiotemporal model with three effects: space, time, and space-time, which uses the travel intensity data during the outbreak of the novel coronavirus (COVID-19) in Hubei province (2020.01.01–2020.05.02). With the help of Markoff’s Monte Carlo method, this paper analyzes the distribution and fluctuation of traffic flow in each city of Hubei province. The results show that the space-time model does not deteriorate compared with the main space model. The study found that nearly 41% of cities with a spatial effect higher than 1 were active during the epidemic in Hubei province and the time effect of travel intensity in Hubei province dropped rapidly from 2 to 0.5 after cities in Hubei province issued measures to close the cities one after another, which lasted nearly a month. Strict social distance intervention is one of the important reasons for Hubei province to control the epidemic effectively in a few months. At the same time, in the stability analysis of the city, we found that Wuhan belongs to an unstable area, which is unfavorable to the control of COVID-19. The research results provide a certain perspective for COVID-19 prevention and control: when there are confirmed patients in the province, we believe that the government should first pay attention to those cities with high spatial effect and instability.

Additive manufacturing, also known as 3D printing or digital manufacturing, has opened up a new era for digital design and fabrication. In recent years, the additive manufacturing industry has been extensively concerned. Currently, cardinal industrial countries in the world are incrementally promoting 3D printing technology as the basis of future manufacturing industry. Additive manufacturing is increasingly employed for mass customization and efficient production in healthcare, automotive industry, electronic industry, and the aerospace industry.

Significantly, miraculous additive manufacturing has excited the aerospace industry. With regard to aerospace applications, additive manufacturing provides unparalleled flexibility in the geometry, material composition, and lead time of components. By means of the manufacturing of highly complex, lightweight, and low-cost components, China is moving towards a revolution in aerospace manufacture. In terms of the current technical situation, especially launch rockets, it is difficult to fabricate the accessories and repairs, structures, and materials of thin-walled aero-engines with complex geometry, which is another fundamental factor that forces aerospace departments to adopt 3D printing technology. From the perspective of sustainable development, additive manufacturing not only has a firm foothold in the domain of aerospace but also becomes the preferred manufacturing tool.

In this paper, the dominating additive manufacturing technologies, the developments of aerospace materials, and their outstanding contributions to the development of the aerospace industry in recent years are summarized. In addition, the superiorities, challenges, and prospects of additive manufacturing in the aerospace industry are also introduced.

Aiming at the common delay phenomenon of maintenance support activities in equipment system, this paper establishes a support delay model based on a Semi-Markov chain, considering the factors of support personnel, support equipment, and spares in the support process. We realize the quantitative description of system instantaneous availability with logistics and maintenance delay. Based on Fourier transform, the convolution equations are solved, while the numerical expression of instantaneous availability varying with time is obtained. By analyzing the variation of instantaneous availability, it is found that support delay has a great impact on system availability. This research enriches the instantaneous availability modeling of multistate system, promoting the solution technology of multistate system in a semi-Markov transition. The direction for reducing instantaneous availability decrease and optimizing system was provided.

Since Google proposed the concept of knowledge graph in 2012, both academic and industrial communities have paid increasing attention to it because of its strong ability of representing information about knowledge and reasoning. Although a lot of studies and applications of knowledge graph have been released in recent years, it is still very few that research related to how to build a knowledge graph involved with Mongolian resources. Therefore, this chapter provided a survey on constructing a high-quality Mongolian vertical domain knowledge graph, and which techniques should be used. First, we introduce the state of the art in the construction of knowledge graph. Second, discuss and analyze the challenges in building a Mongolian domain-specific knowledge graph such as in the financial domain. Finally, we propose an artificial intelligence technique for named entity recognition for Mongolian language resources.

In natural language processing tasks, word segmentation is an important fundamental task, and the accuracy and rationality of Mongolian word segmentation will alleviate the data sparsity problem and directly affect the subsequent work of Mongolian information processing. A BiLSTM-CNN-CRF-based deep learning Mongolian word segmentation model is developed to address the stickiness characteristics of Mongolian and the rich and complex morphological changes of words. The model introduces a convolutional neural network (CNN) algorithm in a bidirectional long- and short-term memory network (BiLSTM) to realize the local features of Mongolian character sequences. The model’s overall performance for Mongolian word segmentation is improved by combining the extraction of local features and long-range time-dependent features and then by using a conditional random field (CRF) model for constrained correction. The LSTM model, BiLSTM model, and BiLSTM-CRF model were selected for comparison experiments. It was verified that the BiLSTM-CNN-CRF-based neural network model performed best on the test data set when the word vector dimension was 100 and the dropout value was 20%, and the introduction of CNN as the local feature extractor of the sequence effectively improved the overall performance of the model. The introduction of CNN as a local feature extractor effectively improved the overall labeling effect of the model, and the F value reached 99.47%.

In order to monitor whether the workers at the construction site are wearing the helmets correctly and ensure their safety, it is necessary to research the method of automatic detection of helmets wearing. This study proposes a method of safety helmet wearing detection based on YOLOv5 algorithm and adds a small target detection layer based on the original multi-scale prediction. The experiments are based on two datasets, which are obtained by open source and self-construct. Through the final verification on the test set, the recognition accuracy of this method reached 92.9% and 90.2%. Comparing with other algorithms, it proves that YOLOv5 has a greater improvement in detection accuracy and speed. Therefore, YOLOv5 algorithm can be deployed to the mobile terminal and realize the transmission of high-definition surveillance video and data collection, real-time monitoring, and warning of behaviors not wearing helmets. It can effectively improve the supervision efficiency of the unsafe behavior at the construction site.