Artificial Internet Of Things Research Articles

Novel Perspectives in Artificial IoT: Current Trends, Research Directions

Artificial Internet of Things (AIoT) amalgamates Artificial Intelligence (AI) technologies with the Internet of Things (IoT), creating a paradigm that promises transformative impacts across various domains. This paper presents an exploration of novel perspectives in AIoT, focusing on current trends and emerging research directions. We begin by providing an overview of the evolving landscape of AIoT, highlighting recent advancements and applications. Subsequently, we delve into the challenges encountered in AIoT systems, encompassing issues such as scalability, security, privacy, and interoperability. The core of this paper lies in delineating promising research directions that offer innovative solutions to address these challenges. We discuss the role of edge and fog computing, security enhancement techniques, AI driven analytics, autonomous systems, human computer interaction, resilient networks, and ethical considerations in shaping the future of AIoT. Through case studies and real world examples, we illustrate the practical implications of embracing novel perspectives in AIoT. Additionally, we forecast emerging trends and provide recommendations for researchers, practitioners, and policymakers to capitalize on the opportunities presented by AIoT. This paper serves as a comprehensive guide for understanding the current state, identifying challenges, and envisioning the future of AIoT research and practice.

Practical Feature Inference Attack in Vertical Federated Learning During Prediction in Artificial Internet of Things

The emergence of edge computing guarantees the combination of the Internet of Things (IoT) and artificial intelligence (AI). The vertical federated learning (VFL) framework, usually deployed by split learning, can analyze and integrate information on different features collected by different terminals in the IoT. The complete model is divided into a top model and multiple bottom models in a specific middle layer. Each passive party as a terminal with certain features owns a bottom model, and an active party as an edge server with labels holds the top model. Feature inference attack aims to infer the party’s features from the model predictions during prediction in VFL. Existing attacks considered the adversary an active party under the white-box or black-box model. However, an attacker usually is a passive party in practice because terminals are more vulnerable than edge servers. Therefore, this paper discusses a practical feature inference attack in VFL during prediction in IoT under this setting. We design an adversary builds an inference model to minimize the distance between the predictions from the inferred features and target features. Because the information on the top model and other bottom models is unknown, the adversary cannot directly train the inference model. Therefore, we utilize the zeroth-order gradient estimation method to calculate the parameters’ gradients to train the inference model. Experimental results demonstrate that the performance of our attack is comparable to that of the white-box attacks while retaining apparent advantages over the existing black-box attacks.

Communication-Efficient and Attack-Resistant Federated Edge Learning With Dataset Distillation

Federated Edge Learning considers a large amount of distributed edge nodes collectively train a global gradient-based model for edge computing in the Artificial Internet of Things, which significantly promotes the development of cloud computing. However, current federated learning algorithms take tens of communication rounds transmitting unwieldy model weights under ideal circumstances and hundreds when data is poorly distributed. This drawback directly results in expensive communication overhead for edge devices. Inspired by recent work on dataset distillation and distributed one-shot learning, we propose Distilled One-Shot Federated Learning (DOSFL) to significantly reduce the communication cost while achieving comparable performance. In just one round, each client distills their private dataset, sends the synthetic data to the server, and collectively trains a global model. The distilled data look like noise and are only useful to the specific model weights, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i.e.,</i> become useless after the model updates. With this weight-less and gradient-less design, the total communication cost of DOSFL is up to three orders of magnitude less than FedAvg while preserving up to 99% performance of centralized training on both vision and language tasks with different models including CNN, LSTM, Transformer, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">etc</i> . We demonstrate that an eavesdropping attacker cannot properly train a good model using the leaked distilled data, without knowing the initial model weights. DOSFL serves as an inexpensive method to quickly converge on a performant pre-trained model with less than 0.1% communication cost of traditional methods.

Mobile Collaborative Secrecy Performance Prediction for Artificial IoT Networks

The integration of artificial intelligence and Internet of Things (IoT) has promoted the rapid development of artificial IoT (AIoT) networks. A wide range of AIoT applications have generated a great deal of data. The fifth-generation (5G) mobile communication has powerful data processing capabilities, and it is a key technology to enable AIoT big data processing. The explosive growth of the 5G users has made information security in AIoT networks a significant issue. Real-time security evaluation in AIoT networks is difficult due to user mobility and dynamic wireless environments. Thus, the evaluation and prediction of secrecy performance is a very critical research. In this article, new expressions for the nonzero secrecy capacity probability (NSCP) are derived to evaluate the mobile collaborative secrecy performance. An improved convolutional neural network (CNN) model, named as SI-CNN in this article, is proposed to predict the NSCP performance. The SI-CNN model combines the SqueezeNet and InceptionNet, and it has four convolution layers, which all adopt the same convolution model. For the first two layers, they employ a 2 × 1 convolution and a three-branch convolution, which not only increase the number of channels but also extract more features. For the last two layers, they employ the same structure, but different convolution kernels. The proposed SI-CNN prediction algorithm is shown to provide better NSCP performance prediction than other state-of-the-art methods. In particular, compared with wavelet neural network, the prediction precision of SI-CNN is improved by 26.8%.

Intelligent camouflage pattern generating in internet of things

Artificial Internet of Things (AIoT) has facilitated many aspects of people's daily lives and greatly promoted the development of military technology. Due to the rapid development of multiple Internet of Things (IoT) devices. The application of AIoT to daily lives and military come true. In our AIoT system, the key part is artificial camouflage pattern generating. In this paper, a new method for generating the camouflage pattern is proposed. In our algorithm, the loss function is designed based on the statistic information of nearby surroundings and traditional camouflage pattern. The statistic information is extracted from the feature maps of convolution neural network. The new camouflage pattern is synthesized by minimize the loss function. A color mapping method is designed. This color mapping method is design for the color mapped from the background to the camouflage pattern. Our method is more flexible than traditional methods and can integrate better with the surrounding environment. Experiments show that this method can effectively generate camouflage patterns similar with background in textures and colors. Besides, our method can well save the texture of traditional artificial designed camouflage pattern.

Deep Reinforcement Learning of Collision-Free Flocking Policies for Multiple Fixed-Wing UAVs Using Local Situation Maps

The evolution of artificial intelligence and Internet of Things (IoT) envision a highly integrated artificial IoT (AIoT) network. Flocking and cooperation with multiple unmanned aerial vehicles (UAVs) are expected to play a vital role in industrial AIoT networks. In this article, we formulate the collision-free flocking problem of fixed-wing UAVs as a Markov decision process and solve it in the deep reinforcement learning (DRL) framework. Our method can deal with a variable number of followers by encoding the dynamic environmental state into a fixed-length embedding tensor. Specifically, each follower constructs a fixed-size local situation map that describes the collision risks with other followers nearby. The local situation maps are used by a proposed DRL algorithm to learn the collision-free flocking behavior. To further improve the learning efficiency, we design a reference-point-based action selection strategy and an adaptive mechanism. We compare the proposed MA2D3QN algorithm with several benchmark DRL algorithms through numerical simulation, and we verify its advantages in learning efficiency and performance. Finally, we demonstrate the scalability and adaptability of MA2D3QN in a semiphysical simulation experiment.

Newton-interpolation-based zk-SNARK for Artificial Internet of Things

Artificial Internet of Things (AIoT) is that the system collects all kinds of information in real-time through various sensors, and intelligence analysis of the data through machine learning in the terminal equipment, edge domains, or cloud centers, including positioning, comparison, forecasting, scheduling, etc. which brings about the data security and privacy issues. The blockchain is a tamper-evident, unforgeable distributed ledger that protects security and privacy through the famous algorithm zk-SNARK, which is also widely used in virtual digital currencies such as Zcash. In addition, by using zk-SNARK technology in the Loopring DEX 3.0 in Ethereum, not only decentralization but also transaction performance can be guaranteed. However, there are three main problems of zk-SNARK, one is the need to guarantee calculation accuracy, two is the long time to generate evidence, especially when using Lagrangian interpolation to QAP the transaction data requires more computation; the last is the poor scalability, especially when nodes need to recalculate all data when adding new transactions. In this paper, we propose a modified zk-SNARK based on Newtonian interpolation, improve the QAP part of zk-SNARK by Newtonian interpolation, and verify the correctness of the scheme through instantiation. Finally, we analyze the computational efficiency of the two interpolation methods, and the results show that Newton interpolation solves the above two problems in the original zk-SNARK, and significantly reduces the time complexity of the algorithm, which can further promote the application of blockchain in data management of AIoT.

Semantic Integration of Sensor Knowledge on Artificial Internet of Things

Artificial Internet of Things (AIoT) integrates Artificial Intelligence (AI) with the Internet of Things (IoT) to create the sensor network that can communicate and process data. To implement the communications and co-operations among intelligent systems on AIoT, it is necessary to annotate sensor data with the semantic meanings to overcome heterogeneity problem among different sensors, which requires the utilization of sensor ontology. Sensor ontology formally models the knowledge on AIoT by defining the concepts, the properties describing a concept, and the relationships between two concepts. Due to human’s subjectivity, a concept in different sensor ontologies could be defined with different terminologies and contexts, yielding the ontology heterogeneity problem. Thus, before using these ontologies, it is necessary to integrate their knowledge by finding the correspondences between their concepts, i.e., the so-called ontology matching. In this work, a novel sensor ontology matching framework is proposed, which aggregates three kinds of Concept Similarity Measures (CSMs) and an alignment extraction approach to determine the sensor ontology alignment. To ensure the quality of the alignments, we further propose a compact Particle Swarm Optimization algorithm (cPSO) to optimize the aggregating weights for the CSMs and a threshold for filtering the alignment. The experiment utilizes the Ontology Alignment Evaluation Initiative (OAEI)’s conference track and two pairs of real sensor ontologies to test cPSO’s performance. The experimental results show that the quality of the alignments obtained by cPSO statistically outperforms other state-of-the-art sensor ontology matching techniques.