In this paper, we address the issue of Channel State Information (CSI) prediction of the Internet of Vehicles (loV) system, which is a highly dynamic network environment. We propose a deep reinforcement learning-based approach to predict CSI with historical data and video footage captured by smart cameras. Specifically, we use a Conventional Neural Network (CNN) to extract unique environmental characteristics, which will be sent to a Recurrent Neural Network (RNN)-based learning model so that the future CSI can be predicted. Our approach also considers the heterogeneous nature of IoV communication environments by adopting transfer learning to reduce the training cost when applying our approach to different IoV scenarios. We assess the efficacy of our proposed approach using our designed IoV simulation platform. The experimental results confirm that our approach can accurately predict CSI by using historically generated data.

Recent years have witnessed the deep involvement of hand gesture recognition in Internet of Things (IoT) devices in healthcare, autonomous driving, virtual reality, augmented reality, etc., since hand gestures offer a natural and intuitive way for human beings to interact with IoT devices without relying on traditional input methods such as keyboards, mice, or touchscreens. Thus, accurate gesture recognition is crucial to its development. Consequently, many recognition approaches are proposed, from the traditional computer vision domain to the deep learning domain. Although with promising recognition performance, these approaches typically come with high computational and energy costs relying on graphics processing unit (GPU) accelerations, which cannot be compatible with low-cost and non-GPU IoT edge devices. To this end, in this paper, we develop an artificial intelligence (AI)-enabled system for real-time hand gesture recognition on low-cost edge devices, e.g., Raspberry Pis. Particularly, we first design a simple but effective convolutional neural network (CNN)-based model with residual blocks to handle American Sign Language (ASL) digits tasks, which enables fast-bust-accurate real-time inferences. Then, to fit the low-cost environment for edge devices, we involve model quantization techniques to shrink the model size, thus requiring reduced memory and storage for edge devices. In addition, we plug a motion sensor into our system to automatically turn off our recognition once it detects no people nearby, reducing energy consumption. By evaluating the performance of real-time hand gesture recognition, our system maintains a high prediction accuracy rate, e.g., over 96 %. Moreover, our designed solution has the potential to be privileged to other common low-cost edge devices.

The Internet of Vehicles (IoV), as one subset of the Internet of Things (IoT) in the smart transportation area, integrates vehicle networks with sensors and actuators. By connecting all sensors to the network, the IoV enables smart transportation (i.e., autonomous vehicles) and makes smart cities a reality. In smart transportation systems, roadside units (RSUs) capture all vehicle information and serve as gateways. However, smart transportation infrastructure has yet to mature in the current stage. RSUs are insufficient to support all vehicles. Meanwhile, the low computational capability of vehicles makes it challenging to recompute the driving route as the road environment changes. To address the problem of insufficient RSU coverage, one protocol called IEEE 802.11p enables vehicle-to-vehicle communication using relays. Nonetheless, data transfer among vehicles via relays is still time-consuming for a large-scale transportation network. To deal with the above issues, in this paper, we propose an IoV framework using digital twins (DTs) to digitize the IoV environment and assign nearby IoT gateways compatible with the RSU communication protocol. This framework lets DTs update the vehicle’s driving route based on real-time information. With a case study, we evaluate the efficacy of DT-assisted IoV based on communication latency and vehicle driving efficiency. Our evaluation results confirm that the proposed framework can efficiently enhance communication latency when the relay needs to pass through two or more vehicles and reduce travel time when vehicles receive updated route information at intersections.

Moving object trajectory tracking plays an important role in traffic scheduling, route planning, advertising recommendations, and other associated social services. The success of moving object trajectory tracking can be attributed to the extensive use of Internet of Things (IoT) devices, which collect a growing volume of spatio-temporal data. To mine the spatio-temporal correlation, traditionally, recurrent neural networks (RNNs) and their variants, such as long-short-term memory (LSTM) and bidirectional long-short-term memory (BiLSTM), have shown their effectiveness in forecasting moving object positions. However, these methods have faced challenges in dealing with complex temporal dependencies due to the limited memory of storing past information using basic hidden layers. To address this issue, in this study, we propose a spatiotemporal attention-based transformer model to mine the spatiotemporal correlation of moving object trajectories in smart transportation systems, which offers improved performance in long-term trajectory prediction tasks. Moreover, most existing works overlook the importance of sampling issues in the trajectory prediction and tracking process. To this end, we develop an adaptive approach by leveraging spatio-temporal sampling to optimize trajectory tracking with reduced data transmission rates and computational costs. The experimental results on real-world datasets demonstrate the superiority of our transformer-based approach over existing RNN-based methods in trajectory predictions and confirm the feasibility of our optimal sampling solution in enhancing trajectory tracking performance.

The Internet of Things (IoT) paradigm is gradually becoming more prevalent through numerous devices and technologies, including sensors, actuators, microcontrollers, cloud-enabled services, and analytics. IoT objects gain intelligence by integrating with wireless sensor networks (WSNs), mobile computing and communication, and others. With sensors, smart things can be enabled by monitoring and identifying environmental changes related to motion, temperature, humidity, pressure, light, vibration, etc. To timely keep track of state changes, researchers are considering developing a cyber replicator, denoted as Digital Twin (DT), of real physical systems as a way to visualize, model, and work with complex cyber-physical systems (CPS). In this paper, we first refine the dataset to a format that can be easily used for deep learning (DL) experiments, IoT data pipeline development, data modeling and simulation, data aggregation, etc. We then demonstrate that DT data can be used to determine space occupancy based on the ambient light sensor, which tends to indicate occupancy in particular spaces because the building has smart lighting that will switch off when rooms are unoccupied after a certain time. Given the apparent developments in machine learning technology, it is clear that machine learning-based prediction has the ability to enhance resource utilization and further forecast future events. Particularly, we use a DT-based dataset and Long-Short-Term Memory (LSTM) neural network architecture to forecast the campus building’s internal temperature.

With the rise and growing attention on Digital Twins (DT) as a way to provide integration between the Internet of Things (IoT) and data analytics, so does the need to consider how to address its challenges. To deal with these challenges, Named Data Networking (NDN) can be a possible solution. NDN has been rising in popularity due to its advancements over the traditional TCP/IP Internet architecture. In this paper, our approach begins with the framework that leverages an NDN-based DT architecture for data management. We then design two scenarios that focus on the performance of data querying in a small and large-scale simulated NDN-based DT architecture. Based on the designed scenarios, we conduct the performance evaluation of data query and DT performance to investigate the performance gap and determine whether an action needs to be taken.

In this paper, we address the issue of automating network configurations for dynamic network environments such as the Internet of Vehicles (IoV). Configuring network settings in IoV environments has proven difficult due to their dynamic and self-organizing nature. To address this issue, we propose a deep reinforcement learning-based approach to configure IoV network settings automatically. Specifically, we use a collection of neural networks to convert the observations of a communication environment (channel power gain, cross-channel power gain, etc.) into key features, which are then supplied to a deep$Q$neural network (DQN) as input for training. Afterward, the DQN will select the optimal network configuration for vehicles in the IoV environment. In addition, our approach considers both centralized and distributed training strategies. The centralized training strategy conducts the DQN training process on a roadside server, while the distributed training strategy trains the DQN on vehicles locally. Through our designed IoV simulation platform, we evaluate the efficacy of our proposed approach, demonstrating that it can improve the quality of services (QoS) in the IoV environments concerning reliability, latency, and service satisfaction.

The interconnection and digitization of the physical world has increased dramatically with the widespread deployment of network communication and the rapid development of the Internet of Things (IoT). Application scenarios and requirements in IoT are more complex and diverse than ever before. To successfully support the design and development of complex IoT systems, a realistic evaluation platform that can accurately simulate both the physical world and network communications is necessary. Yet, most existing simulation tools are limited, simulating only specific subsets of IoT environments, such as communication network simulation or mobility simulation, rather than complete IoT scenarios. Thus, in this paper, we propose a new framework, in which several modules can work together to achieve more realistic simulation of IoT environments. Specifically, we integrate three-dimensional object motion with the OMNET++ network simulator. In our framework, we can configure and direct object movement in 3D and compute the received power of transmitted signals using ray tracing techniques. Within the framework, OMNET++ simulates the communication process based on the received power and communication protocol. As a demonstration of our framework, we conduct several experiments on two classic Internet of Vehicles (IoV) scenarios. The results indicate that our proposed framework can accurately simulate both the physical and communication aspects of IoT systems.

The Internet of Things (IoT) encompasses a near-incalculable collection of dispersed and embedded computing devices acting as sensors and actuators, generating data at an incredible scale. However, a lack of coherency and cross-compatibility in IoT deployments has lead to increasing redundancy and waste of resources. To combat this, various concepts have been proposed for an open IoT search engine (IoT-SE) that serves human and machine users. Invariably, the IoT-SE envisions distributed query retrieval to handle massive volumes of devices and data. Incorporating the massively heterogeneous protocols and properties of devices deployed, the search of such a system for timely and pertinent data is massively challenging, to provide useful knowledge and service for IoT systems. Moreover, enabling and maintaining security and privacy in an IoT-SE is likewise a prodigious task, as end users, IoT devices, and the search system itself, have different protocols and requirements. To this end, a study of security issues in IoT search is conducted to outline the challenges ahead, and a case study to resolve practical security vulnerabilities in an IoT-SE system is carried out. The pertinent issues of security in an IoT-SE system are reviewed. Particularly: 1) a taxonomy is detailed for IoT-SE security issues; 2) the vulnerabilities of machine learning (ML) models in the IoT-SE are considered; and 3) defensive mechanisms are presented for securing IoT Search. A case study is carried out to implement basic security features in the IoT search, addressing the risk of false queries through the design of ML-based solutions. Finally, a roadmap for future research is provided, including the security and privacy for IoT systems connected to the IoT-SE, distributed edge computing in IoT-SE, privacy-preserving data markets in IoT-SE, and distributed ML in IoT-SE.

Machine learning, as a viable way of conducting data analytics, has been successfully applied to a number of areas. Nonetheless, the lack of sufficient data is one critical issue for applying machine learning in Industrial Internet of Things (IIoT) systems. Insufficient data raises could negatively affect the accuracy of machine learning models. To tackle this issue, we design a framework to systematically investigate the impacts of insufficient data on model training. This framework employs the generative adversarial network (GAN) and continuous learning to generate and engage new data in model training, enabling us to study the security risks of introducing new data in the model training process and develop countermeasures to mitigate these risks. To validate the efficacy of our framework, we consider a representative IIoT scenario, in which a variety of industrial components needs to be recognized by convolutional neural networks (CNNs), and design and implement three evaluation scenarios that are based on a real-world IIoT data set. Our experimental results confirm that insufficient data can have a significant impact on the model accuracy, but that new data generated by GAN and continuous learning can greatly improve the model accuracy. Our experimental results also show that the data poisoning threat posed by the GAN can significantly reduce the model accuracy. However, our proposed defensive mechanism is capable of securing the model learning process. We conclude this article by discussing some emerging issues that need to be addressed in future work.