Article
Automation & Control Systems
Weifeng Gao, Zhiwei Zhao, Geyong Min, Qiang Ni, Yuhong Jiang
Summary: This article proposes a resource allocation scheme RaFed for FL and demonstrates through experiments that it significantly reduces training latency in IIoT systems.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Engineering, Electrical & Electronic
Jinke Ren, Guanding Yu, Guangyao Ding
Summary: Training tasks in classical machine learning models are usually performed at remote cloud centers, which can be time-consuming and resource-heavy, posing privacy and communication latency issues. To address this, federated edge learning framework aggregates local learning updates at network edge, aiming to accelerate the training process.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2021)
Article
Computer Science, Information Systems
Hao Chen, Shaocheng Huang, Deyou Zhang, Ming Xiao, Mikael Skoglund, H. Vincent Poor
Summary: This study investigates the problem of jointly optimizing communication efficiency and resources for federated learning over wireless Internet of Things (IoT) networks. By proposing a new client scheduling policy and a power and bandwidth allocation method, this research improves both the communication efficiency and learning performance.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Computer Science, Information Systems
Van-Dinh Nguyen, Shree Krishna Sharma, Thang X. Vu, Symeon Chatzinotas, Bjorn Ottersten
Summary: Federated learning allows edge computing nodes to build a shared learning model without transferring raw data, but faces challenges like non-iid data and device heterogeneity. To address these challenges and reduce communication overhead, a new FL algorithm is proposed and applied in wireless IoT networks. Experimental results show improved convergence rate and energy efficiency with the new algorithm.
IEEE INTERNET OF THINGS JOURNAL
(2021)
Article
Engineering, Electrical & Electronic
Tra Huong Thi Le, Nguyen H. Tran, Yan Kyaw Tun, Minh N. H. Nguyen, Shashi Raj Pandey, Zhu Han, Choong Seon Hong
Summary: This paper investigates the incentive mechanism design between the base station and mobile users in federated learning through an auction game. The proposed greedy auction mechanism can guarantee truthfulness, individual rationality, and efficiency.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2021)
Article
Engineering, Electrical & Electronic
Xiumei Deng, Jun Li, Chuan Ma, Kang Wei, Long Shi, Ming Ding, Wen Chen, H. Vincent Poor
Summary: This article proposes a novel BFL framework that integrates training and mining at the client side, aiming to optimize the learning performance and address the vulnerability of decentralized model aggregation in existing BFL frameworks.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2023)
Article
Computer Science, Hardware & Architecture
Yuan-Ai Xie, Jiawen Kang, Dusit Niyato, Nguyen Thi Thanh Van, Nguyen Cong Luong, Zhixin Liu, Han Yu
Summary: Federated Learning Networks (FLNs) enable collaborative training of models among mobile devices without compromising local privacy. However, FLNs are vulnerable to various attacks due to frequent model updates. In this article, we propose the Covert Communication-based Federated Learning (CCFL) approach to enhance the privacy and security of FLNs by utilizing covert communication techniques. Experimental evaluation demonstrates the effectiveness of CCFL in terms of training efficiency and communication security under real-world settings.
Article
Engineering, Electrical & Electronic
Yanmeng Wang, Yanqing Xu, Qingjiang Shi, Tsung-Hui Chang
Summary: Federated learning is a viable distributed learning paradigm that collaboratively trains machine learning models on massive mobile devices at the wireless edge while protecting user privacy. However, challenges such as transmission outage and quantization errors can severely jeopardize the FL convergence, prompting the proposal of a robust FL scheme named FedTOE to address these issues and improve performance.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2022)
Article
Telecommunications
Ticao Zhang, Shiwen Mao
Summary: Federated learning is a new approach for resource-intensive and privacy-aware learning applications, allowing collaborative model training without sharing private data. This research proposes leveraging intelligent reflecting surfaces (IRS) to reconfigure the wireless propagation environment and maximize resource utilization, specifically targeting energy efficiency in a reconfigurable wireless communication network. Simulation results confirm the algorithm's fast convergence and significant energy savings, particularly with a large number of reflecting elements and proper IRS configuration.
IEEE TRANSACTIONS ON GREEN COMMUNICATIONS AND NETWORKING
(2022)
Article
Computer Science, Information Systems
Zhanpeng Yang, Yuanming Shi, Yong Zhou, Zixin Wang, Kai Yang
Summary: This article proposes a decentralized blockchain-based federated learning architecture, which uses a secure global aggregation algorithm to resist malicious devices and a practical Byzantine fault tolerance consensus protocol to prevent model tampering from the malicious server. It also formulates a network optimization problem and leverages deep reinforcement learning algorithm to reduce training latency.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Computer Science, Information Systems
Wanli Ni, Yuanwei Liu, Zhaohui Yang, Hui Tian, Xuemin Shen
Summary: This article investigates the use of multiple reconfigurable intelligent surfaces (RISs) to address the problem of model aggregation in federated learning systems. The seamless integration of communication and computation is achieved through over-the-air computation (AirComp), and the mean-square error (MSE) and device set in the model uploading process are optimized to improve the accuracy and convergence rate of federated learning.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Engineering, Electrical & Electronic
Ning Huang, Minghui Dai, Yuan Wu, Tony Q. S. Quek, Xuemin Shen
Summary: This paper proposes a FL framework with a hybrid centralized training and local training to address the coexistence problem of privacy-sensitive and resource-constrained client-devices. Experimental results demonstrate the advantages of the proposed framework and algorithm compared to other FL schemes and algorithms.
IEEE JOURNAL OF SELECTED TOPICS IN SIGNAL PROCESSING
(2023)
Article
Computer Science, Hardware & Architecture
Tinghao Zhang, Kwok-Yan Lam, Jun Zhao, Feng Li, Huimei Han, Norziana Jamil
Summary: This paper proposes a spectrum allocation optimization mechanism for enhancing Federated Learning over a wireless mobile network, aiming to minimize the time delay of Federated Learning while considering the energy consumption of participating devices, ensuring sufficient resources for all devices to train their local models. Additionally, a robust device selection method is proposed to enable rapid convergence of Federated Learning on non-iid datasets.
IEEE-ACM TRANSACTIONS ON NETWORKING
(2023)
Article
Engineering, Electrical & Electronic
Wanli Wen, Zihan Chen, Howard H. Yang, Wenchao Xia, Tony Q. S. Quek
Summary: This paper proposes a training algorithm for hierarchical federated edge learning (H-FEEL) system, which achieves helper scheduling and communication resource allocation through phases like local gradient computing, weighted gradient uploading, and model updating. By mathematical modeling and analyzing convergence bounds, the optimization problem considering wireless channel uncertainty and weighted gradient importance is solved. The effectiveness of the scheme is demonstrated through simulations.
IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS
(2022)
Article
Computer Science, Information Systems
Rafael Valente da Silva, Jinho Choi, Jihong Park, Glauber Brante, Richard Demo Souza
Summary: This paper proposes an optimal method for allocating wireless resources in a multi-channel ALOHA setup, which outperforms uniform and fully-shared channel allocations in terms of convergence time in large-scale wireless sensor networks involving decentralized machine learning.
IEEE WIRELESS COMMUNICATIONS LETTERS
(2022)
Review
Engineering, Environmental
Ye Chen, Shilong Li, Shiru Lin, Mingzhe Chen, Cheng Tang, Xinghui Liu
Summary: Graphite-based materials have gained significant attention and rapid development in the ore industry due to their structure and excellent conductivity. However, there is still a lack of a complete industrial chain from raw materials to commercial products. To accelerate the industrialization process, optimizing the beneficiation process is crucial for improving the valuable minerals by addressing the differences in their physical and chemical properties. This review focuses on the standard beneficiation methods and presents a comprehensive flotation process for high-quality graphite flake, and also discusses the energy storage applications of graphite-based materials.
CHEMICAL ENGINEERING JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Zhaohui Yang, Mingzhe Chen, Zhaoyang Zhang, Chongwen Huang
Summary: In this paper, the problem of wireless resource allocation and semantic information extraction for energy efficient semantic communications over wireless networks with rate splitting is investigated. A base station (BS) first extracts semantic information from its large-scale data, and then transmits the small-sized semantic information to each user which recovers the original data based on its local common knowledge. The joint computation and communication problem is formulated as an optimization problem aiming to minimize the total communication and computation energy consumption of the network under computation, latency, and transmit power constraints. An alternating algorithm is proposed to solve this problem, and numerical results demonstrate the effectiveness of the proposed algorithm.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2023)
Article
Computer Science, Information Systems
Ali Pourranjbar, Georges Kaddoum, Walid Saad
Summary: Conventional anti-jamming methods are ineffective against a single jammer following multiple different jamming policies or multiple jammers with distinct policies. This article proposes an anti-jamming method that can adapt to the current jamming attack and estimates future occupied channels in the multiple jammers scenario. The proposed methods outperform the baseline method and achieve high success rates and ergodic rates.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Review
Computer Science, Information Systems
Guangxu Zhu, Zhonghao Lyu, Xiang Jiao, Peixi Liu, Mingzhe Chen, Jie Xu, Shuguang Cui, Ping Zhang
Summary: Pushing artificial intelligence (AI) from central cloud to network edge has gained consensus in both industry and academia for realizing the vision of artificial intelligence of things (AIoT) in the sixth-generation (6G) era. This has spurred the emergence of edge intelligence, which focuses on extracting human-like intelligence from the vast amount of data at the wireless network edge. This study provides an overview of seamlessly integrated sensing, communication, and computation (ISCC) for edge intelligence, discussing its concept, design challenges, enabling techniques, state-of-the-art advancements, and future prospects.
SCIENCE CHINA-INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Sheikh Salman Hassan, Do Hyeon Kim, Yan Kyaw Tun, Nguyen H. H. Tran, Walid Saad, Choong Seon Hong
Summary: This study investigates the design of an energy-efficient resource allocation system for non-terrestrial networks (NTNs) that integrates space and aerial networks with terrestrial systems. The goal is to maximize system energy efficiency by optimizing user equipment association, power control, and unmanned aerial vehicle deployment. The study proposes a mixed-integer nonlinear programming problem and develops an algorithm to decompose and solve each problem distributedly. Simulation results demonstrate that the algorithm achieves better energy efficiency and spectral efficiency than baselines.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Editorial Material
Engineering, Multidisciplinary
Zhi Zhou, Dusit Niyato, Zehui Xiong, Xiaowen Gong, Walid Saad, Xiaoming Fu
Summary: The papers in this special issue discuss the interaction between edge computing and artificial intelligence (AI) in 6G mobile communication networks. They focus on the potential of 6G networks to create an Internet of Intelligence by connecting people, things, and intelligence to solve human challenges and improve our world. Edge computing, which pushes computing tasks and services from the network core to the edge, is recognized as an essential component for empowering 6G networks with AI capabilities.
IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING
(2023)
Article
Telecommunications
Kitae Kim, Yan Kyaw Tun, Md. Shirajum Munir, Walid Saad, Choong Seon Hong
Summary: Accurate channel estimation and allocation are crucial for delivering RIS-aided wireless network services. A new learning-based method is proposed for channel estimation in RIS systems, which uses a limited number of pilot signals trained in a masked autoencoder (MAE) to achieve high accuracy. Furthermore, a deep reinforcement learning (DRL) agent learns pilot allocation policies through MAE, resulting in higher channel estimation performance with fewer pilots than conventional methods.
IEEE COMMUNICATIONS LETTERS
(2023)
Article
Engineering, Electrical & Electronic
Lunan Sun, Yang Yang, Mingzhe Chen, Caili Guo, Walid Saad, H. Vincent Poor
Summary: This paper proposes an adaptive information bottleneck guided joint source and channel coding (AIB-JSCC) method for image transmission, aiming to reduce the transmission rate while improving the image reconstruction quality. By dynamically adjusting the hyperparameter of the loss function, a balance between compression and reconstruction quality is achieved. Experimental results demonstrate that AIB-JSCC can significantly reduce the amount of transmitted data and improve the reconstruction quality and downstream task accuracy.
IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Do Hyeon Kim, Aunas Manzoor, Madyan Alsenwi, Yan Kyaw Tun, Walid Saad, Choong Seon Hong
Summary: This paper proposes a novel framework for analyzing data offloading in a multi-access edge computing system, which includes an algorithm with two key phases. The first phase utilizes ruin theory to handle user association, taking into account their transmission reliability and resource utilization efficiency. The second phase employs an optimization-based algorithm to optimize the data offloading process and minimize users' energy consumption.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Article
Engineering, Electrical & Electronic
Zhuojun Tian, Zhaoyang Zhang, Zhaohui Yang, Richeng Jin, Huaiyu Dai
Summary: In conventional distributed learning over a network, the unified learning model becomes inefficient for each agent due to the underlying non-i.i.d. data distribution among agents. To address this problem, we propose a graph-attention-based personalized training algorithm (GATTA) for distributed deep learning.
IEEE TRANSACTIONS ON SIGNAL PROCESSING
(2023)
Article
Telecommunications
Yihan Cang, Ming Chen, Jingwen Zhao, Tantao Gong, Jiahui Zhao, Zhaohui Yang
Summary: This paper investigates the joint optimization of task offloading and resource allocation for OFDMA-based multi-access edge computing systems. An iterative algorithm is proposed to solve the non-convex mixed-integer problem, and simulation results show the superiority of the proposed algorithm in terms of minimum computation efficiency.
IEEE COMMUNICATIONS LETTERS
(2023)
Article
Chemistry, Multidisciplinary
Mohammad Fahda, Manal Ammar, Walid Saad, Mohamad Hmadeh, Mazen Al-Ghoul
Summary: In this study, the high-quality synthesis of ZIF-(8, 67) crystals and their mixed metal derivatives was achieved in an aqueous medium with reduced organic ligand consumption and controlled particle size. The rapid and precise control over the transition from ZIF-L to ZIF-(8, 67) was demonstrated using flow rates and molar ratios of the initial ZIF precursors. The synthesis of smaller ZIF-8 nanoparticles and the controlled doping of ZIF-8 with cobalt were also achieved.
Article
Computer Science, Information Systems
Chunyu Pan, Jincheng Wang, Xinwei Yue, Linyan Guo, Zhaohui Yang
Summary: This paper proposes an urban-micro CF-UAV (UMCF-UAV) network architecture and a dynamic resource allocation algorithm to reduce system delay in UAV-assisted cellular networks. Simulation results show that the proposed algorithm has fast convergence behavior and achieves decreased system delay compared to other baseline resource allocation schemes, with a maximum improvement of 53%.
Article
Computer Science, Information Systems
Yantong Wang, Ye Hu, Zhaohui Yang, Walid Saad, Kai-Kit Wong, Vasilis Friderikos
Summary: This paper proposes a novel framework for proactive caching that combines model-based optimization with data-driven techniques. It transforms an optimization problem into a grayscale image and uses Convolutional Neural Networks (CNNs) to predict content caching locations. Two algorithms are provided to address competition and accelerate search.
IEEE TRANSACTIONS ON MOBILE COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Minh N. H. Nguyen, Shashi Raj Pandey, Tri Nguyen Dang, Eui-Nam Huh, Nguyen H. Tran, Walid Saad, Choong Seon Hong
Summary: This article introduces a distributed learning approach based on the philosophy of democratized learning (Dem-AI), which includes a self-organizing hierarchical structuring mechanism and solutions for hierarchical generalized learning problems. Experimental results show that the proposed algorithm outperforms conventional federated learning algorithms in the generalization performance of learning models in agents.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)