Internet of Things最新文献

Integrating Artificial Intelligence (AI) with the Internet of Things (IoT) has propelled technological innovation across various industries. This systematic literature review explores the current state and future trajectories of AI in IoT, with a particular focus on emerging trends in intelligent data analysis and privacy protection. The proliferation of IoT devices, marked by voluminous data generation, has reshaped data processing methods, providing actionable insights for informed decision-making. While previous reviews have offered valuable insights, they often must comprehensively address the multifaceted dimensions of the AI-driven IoT landscape. This review aims to bridge this gap by systematically examining existing literature and acknowledging the limitations of past studies. The study uses a meticulous approach guided by established methodologies to achieve this aim. The chosen methodology ensures the rigour and validity of the review, aligning with PRISMA 2020 guidelines for systematic reviews. This systematic literature review serves as a comprehensive guide for researchers, practitioners, and policymakers, offering insights into the current landscape and paving the way for future research directions. The identified trends and challenges provide a valuable resource for navigating the evolving domain of AI in IoT, fostering a balanced, secure, and sustainable advancement in this dynamic field. Our analysis shows that integrating AI with IoT improves operational efficiency, service personalisation, and data-driven decisions in healthcare, manufacturing, and urban resource management. Real-time machine learning algorithms and edge computing solutions are set to revolutionise IoT data processing and analysis by improving system responsiveness and privacy. However, increasing concerns about data privacy and security emphasise the need for new regulatory frameworks and data protection technologies to ensure the ethical adoption of AI-driven IoT technologies.

Today is the age of superconductivity where each object connects in a cascading manner to other objects, allowing for seamless integration of real-world objects into the digital domain of the Internet of Things (IoT). These objects collaborate to deliver ubiquitous services based on the user mode and context. For more real-time applications, IoT is integrated with quantum computing technologies and tools for enhancing the conventional structure into more different aspects, revolutionizing the processing speed, enhancing communication, and increasing security features. All these objects are equipped with sensors that collect real-time data from their surroundings and share it with neighboring objects. This data is then broadcast into the environment, enabling users to access services without understanding the underlying complex and hybrid IoT infrastructure of heterogeneous devices. These minute and plugable sensors are capable of data collection and are always busy handling data management. However, these sensors often have limited resources, creating significant issues when dealing with massive and repetitive operations. Most of the time, these low-energy sensors are busy with excessive sensing and broadcasting, resulting in overhearing and passive listening. These factors not only create congestion on communication channels but also increase delays in data transmission and adversely affect system performance. To assess the network traffic for securing the IoT resources in the quantum computing environment, in this research work, we have proposed a novel scheme called “Self-Adaptive and Content-Based Scheduling (CACS) for Reducing Idle Listening and Overhearing in Securing the Quantum IoT Sensors”. This scheme reduces idle listening and minimizes overhearing by adaptively configuring network conditions according to the contents of sensed data packets. It minimizes extensive sensing, decreases over-cost processing, and reduces frequent communication that lessens the overall system traffic and secures the resources from being overwhelmed. The simulation results demonstrate a 0.80% increase in delay across various baud rates, resulting in a general increase of 0.44 s. Moreover, it ensures a notable 22.23% reduction in BER and lowers energy consumption by approximately 20%, which is actual energy enhancement in the connected system.

In recent times, a new paradigm has emerged in the field of Cloud computing, namely Fog computing. This paradigm has proven to be highly useful in a wide range of domains where both delay and cost were important metrics. Notably, the Internet of Things (IoT) strongly benefits from this, as small devices can gain access to strong computation power quickly and at a low cost. To achieve this, task offloading is used to decide which task should be executed on which node. The development of an efficient algorithm to address this problem could significantly enhance the sustainability of systems in various industrial, agricultural, autonomous vehicle, and other domains. This paper proposes a new variant of the Niche Pareto Genetic Algorithm (NPGA) called Local search Drafting-NPGA (LD-NPGA) to optimize resource allocation in a Cloud/Fog environment, with the objective of minimizing makespan and cost simultaneously. It generates Pareto solutions allowing the user to make choices closer to its intentions. Thus, it addresses various shortcomings identified in the state of the art, including scalability and aggregation formula. A drafting step is implemented to maintain diversity in the population of solutions, resulting in a more varied Pareto set than basic NPGA. LD-NPGA significantly outperforms state-of-the-art metaheuristics in makespan and cost by 15%. Finally, the scalability of our approach and the variety of solutions generated are confirmed in the different experiments.

This paper presents object detection methods to accurately identify the sources of flame and smoke in vast circumstances. Aerial drones collected the data, analyzed the recognition outputs in real time on an edge device, and then transferred them to the back-end for data processing and warnings using Kafka. To detect flame and smoke occurrences, the models were compared using various convolutional neural networks (CNN). Several factors considered include streaming speed, accuracy, portability, efficiency, and power consumption on edge devices. This work conducted training comparisons of YOLOV4, YOLOV5, YOLOV7, YOLOV8, and Faster RCNN. The inference performance was then evaluated on an edge computing device. The findings showed an accuracy of 0.91 and 0.87, while maintaining a processing speed of roughly 1 frame per second on the Nvidia Jetson NX without acceleration.

Autonomous vehicles have attracted attention as a result of enhancements in artificial intelligence, the Internet of Things, and communication technologies. There is a priority for new testing frameworks to keep up with the increasing complexity of decision-making, connectivity, data interchange, and data transmission speed. Traditional vehicle testing tools and methods cannot meet the new testing requirements imposed by the upgrading of autonomous driving (AD) technology. They are expensive, time-intensive, and present safety hazards during testing, as well as cannot simulate hybrid real-world situations and manage real-time data efficiently. Therefore, in terms of test efficiency, cost, and safety, a smart vehicle testing in hybrid reality and evaluation method based on a digital twin (DT) is presented to speed up the development and testing of AD functions. Our model uses three-dimensional coordinate mapping, a collision detection model, and virtual scene registration to plot the AD information in the actual environment to the virtual scenario. In addition, the consistent mixed reality-based AD test model is constructed at the same time. Using this model, we demonstrated that our proposal allows for better performance of an AD test. Furthermore, the collision test demonstrates that the mixed reality system has interactive features. The performance of the system under the sampling rate of 50 ms, 200 ms, and 800 ms is compared and analyzed. Also, the experiments show that the algorithm described in this paper works better when the sampling frequency is 200 ms or more.

This paper investigates joint beamforming and artificial noise (AN) design for secure multiple-input single-output (MISO) ultra-reliable and low latency communication (URLLC) networks in Internet of Things (IoT) applications. In considered system, a base station (BS) transmits confidential information for individual IoT users using the short-packet communication technique under the wiretap of eavesdroppers. To enhance physical layer security, the BS injects additional dedicated AN symbols to degrade the information retrieval ability at eavesdroppers. In this paper, we aim to joint design the transmit beamforming and AN symbol to maximize the minimum URLLC secrecy rate of IoT user subject to the power budget of the BS. The optimization design problem is highly nonconvex due to coupled variables in the URLLC secrecy rate and channel dispersion expressions, and thus it is mathematically challenging to solve this problem directly. To overcome this issue, we first introduce various convex inner approximations to convexify the nonconvex terms, and then we develop an efficient iterative algorithm based on the sequential convex programming approach. Extensive numerical simulation results are conducted to investigate the URLLC secrecy rate region. In conclusion, the two new URLLC parameters, i.e., the transmit packet blocklength and block error probability, will cause the considerable degradation on the URLLC secrecy rate region when comparing to that of the traditional beamforming design based on the Shannon capacity.

Fog node selection is a crucial element in the development of a fog computing system. It forms the foundation for other techniques such as resource allocation, task delegation, load balancing, and service placement. Fog consumers have the task of choosing the most suitable and reliable fog node(s) from the available options, based on specific criteria. The study presents the Fog Node Selection Engine (FNSE) as an intelligent and reliable fog node selection framework to select appropriate and reliable fog nodes in a trustworthy manner. The FNSE predicts the trust value of fog nodes to help the fog consumer select a reliable fog node based on its trust value. We propose three AI-driven models within the FNSE framework: FNSE based on fuzzy logic (FL), FNSE based on logistic regression (LR), and FNSE based on a deep neural network (DNN). We implement these three models separately using MATLAB for FL and Python for LR and DNN. The performance of the proposed models is compared based on the performance metrics of accuracy, precision, recall, F1 score and execution time. The experiment results show that the FL-based FNSE approach achieves the best performance with the highest accuracy, precision, recall, and F1 score values. The FL-based FNSE approach also consumes less time and can make predictions quickly. The FNSE framework based on FL improves the overall performance of the selection process of fog nodes.

At the forefront of automobile safety technology, Automatic Emergency Braking (AEB) represents a major advancement in collision avoidance systems. The cutting-edge technology provides an additional layer of security at pivotal times, making it a vital part of the changing landscape of vehicle safety. This research introduces an effective system for automatic emergency braking in ADAS-equipped or Autonomous vehicles using a combination of 3d lidar and stereo vision camera for a swift and robust system in the vehicle that is faster than human drivers in cases of unexpected emergencies. Utilizing the power of clustering algorithms on 3d point clouds and state-of-the-art computer vision algorithms on an RGB image mapped to a depth frame from the stereo vision camera, the system as a whole provides a comprehensive system adding to the safety of the vehicle and the passengers. Further, the efficiency of the system is studied based on various parameters. The data from an external inertial measurement unit is also utilized to derive results that support the claims of the study. The system has been developed and implemented on a passenger car that has been modified into an electric vehicle and further tested in real-world traffic conditions in autonomous driving mode. The findings of the study were having exceptionally good precision in split-second decision-making in emergency maneuvers.

Understanding your customers is among one of the most important strategies to boost retail profit. In this research, we propose a WiFi-based sensing method to analyze customer behaviors. The monitoring of customer behaviors may lead to revenue growth. Specifically, the strategy is focused on understanding and grouping customers’ behaviors in which we track customers who share similar visiting patterns through WiFi sensing. Accordingly, we can have group-based prediction done for customers who own similar behaviors. We extract customers’ visiting patterns including the customers’ Service Set Identifier list and related information. After all, the proposed system is realized in a cafeteria place where we have the deployed WiFi access points continuously collect data over a time horizon of three months to serve as the inputs for data analysis. The data samples include the number of customers’ devices, number of products and revenue amounts. The dataset also integrates group information and weather conditions. We adopt several machine learning methods including Support Vector Regression and Random Forest for model induction. We conduct these models in terms of three main prediction tasks consisting of coffee shop’s revenue, the number of products, and the number of customers’ devices for evaluation. Furthermore, considering these predictions, we separate between the staying-in and to-go parts. Based on the experiment result, customers’ group information helps, as well as weather conditions. Overall, we can achieve the best prediction result when both the group information and weather conditions are included where we can enjoy as good as 6% to 10% in MAPE.

In IoT infrastructure, high-frequency sensing and subsequent transmission of sensed data to computational facilities can lead to redundant data storage and processing, consuming significant storage and processing capacity. As a result, the IoT infrastructure needs more data transmission cycles, leading to data redundancy and low network up-time due to the drainage of limited battery capacity. Conversely, if the data is communicated at a lower rate, it may cause absolute data delivery to the processing unit, which is useless. As a result, a well-designed data aggregation algorithm is required. This paper proposes the $AD{A}^2-IoT$, an Adaptive Data Aggregation Algorithm for IoT Infrastructure tailored to optimize parameters such as low data redundancy, limited data communication cycles, and high IoT infrastructure up times. The proposed algorithm consists of two key components: the Route Data Aggregator (RDA) performs aggregation during data transit towards the Edge node or gateway, and the Node Data Aggregator (NDA) performs data aggregation during capturing or sensing data. The algorithm employs metrics like Age of Information (AoI) and data freshness factor during the node and route data aggregation phase to capture and timely deliver data to the Edge node, where this data is processed for informed decision-making. The proposed algorithm was efficiently tested on a simulation and IoT hardware deployment environment. Both simulation and hardware results demonstrate a substantial improvement in QoS parameters, such as a decrease in data redundancy and packet exchanges, leading to considerable energy savings and prolonging the lifespan of IoT infrastructure.