Federated Learning via Over-the-Air Computation
Open accessThe DOI matches its Crossref record; 61 of the 62 references it cites were found in a registry or an index.
Abstract
The stringent requirements for low-latency and privacy of the emerging high-stake applications with intelligent devices such as drones and smart vehicles make the cloud computing inapplicable in these scenarios. Instead, edge machine learning becomes increasingly attractive for performing training and inference directly at network edges without sending data to a centralized data center. This stimulates a nascent field termed as federated learning for training a machine learning model on computation, storage, energy and bandwidth limited mobile devices in a distributed manner. To preserve data privacy and address the issues of unbalanced and non-IID data points across different devices, the federated averaging algorithm has been proposed for global model aggregation by computing the weighted average of locally updated model at each selected device. However, the limited communication bandwidth becomes the main bottleneck for aggregating the locally computed updates. We thus propose a novel over-the-air computation based approach for fast global model aggregation via exploring the superposition property of a wireless multiple-access channel. This is achieved by joint device selection and beamforming design, which is modeled as a sparse and low-rank optimization problem to support efficient algorithms design. To achieve this goal, we provide a difference-of-convex-functions (DC) representation for the sparse and low-rank function to enhance sparsity and accurately detect the fixe…
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“For instance, over-the-air federated learning (FL) provides a collaborative ML framework to train a global statistical model over wireless networks without accessing edge devices' private raw data [31].”
From Edge artificial intelligence for 6g: vision, enabling technologies, and applications (2022) · cites this paper as [31]
In the citing paper
“Specifically, the transmitted signals at edge devices are first multiplied by the fading channels and then superposed over-the-air with additive channel noise, resulting in a noisy weighted sum of transmitted signals [31].”
From Edge artificial intelligence for 6g: vision, enabling technologies, and applications (2022) · cites this paper as [31]
In the citing paper
“Specifically, for edge training systems, we need to jointly optimize the subcarrier and bandwidth allocation [88], [90], [241], transmit power and receive beamforming [31], [48], [178], passive beamforming at RIS [48], [178], device selection [31], [242] and activity detection [76], local updates computation [92], and global aggregation frequency control [130], thereby reducing the optimality gap”
From Edge artificial intelligence for 6g: vision, enabling technologies, and applications (2022) · cites this paper as [31]
In the citing paper
“For instance, the group sparsity can represent the combinatorial variables for edge devices selection in FL [31], edge devices activity detection [76], and inference tasks selection [53].”
From Edge artificial intelligence for 6g: vision, enabling technologies, and applications (2022) · cites this paper as [31]
In the citing paper
“Most of the resource allocation problems in edge AI need to solve a series of nonconvex optimization problems, e.g., nonconvex sparse optimization for device selection in wireless FL, nonconvex quadratic programming for transceiver design in over-the-air FL [31], lowrank matrix optimization for interference management in edge device distributed inference [52], and unit modulus constrained phase”
From Edge artificial intelligence for 6g: vision, enabling technologies, and applications (2022) · cites this paper as [31]
2 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Hao, X.-B., & Zhang, G.-G. (2017) · Encycl. Semantic Comput. Robotic Intell. · doi:10.1142/s2425038416300184 · Cited by 86,401 · cited 1×
“THE astounding growth in data volume promotes widespread artificial intelligent applications such as image recognition and natural language processing [1], thanks to”
BackgroundLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Boyd, S., & Vandenberghe, L. (2004) · doi:10.1017/cbo9780511804441 · Cited by 56,601
Vorontsova, E., Hildebrand, R., Gasnikov, A. V., & Stonyakin, F. (2010) · Computer Vision · doi:10.1007/978-1-4614-5257-7_6 · Cited by 39,270 · cited 2×
“The resulting problem is known as the sum-of-infeasibilities in the literature of optimization [39].”
Learning Multiple Layers of Features from Tiny Images
Krizhevsky, A. (2024) · doi:10.57702/zp44cu3g · Cited by 23,395
Communication-Efficient Learning of Deep Networks from Decentralized Data
McMahan, H. B., Moore, E., Ramage, D., Hampson, S., & Arcas, B. A. Y. (2016) · arXiv · doi:10.48550/arxiv.1602.05629 · Cited by 14,215 · cited 7×
“In particular, the federated averaging (FedAvg) algorithm [12] turns out to be a promising way to efficiently average the locally updated model at each device with unbalanced and non-IID data, thereby reducing the number of communication rounds between the center node and the end devices.”
InfluentialFederated Machine Learning: Concept and Applications
Yang, Q., Liu, Y., Chen, T., & Tong, Y. (2019) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/3298981 · Cited by 6,322
Han, S., Mao, H., & Dally, W. J. (2015) · arXiv · doi:10.48550/arxiv.1510.00149 · Cited by 4,662 · cited 1×
“To address this issue, there is a growing body of recent works to reduce the storage overhead, time and power consumption in the inference process using the model compression methods via hardware and software co-design [7], [8].”
MethodTowards Federated Learning at Scale: System Design
Bonawitz, K., Eichner, H., Grieskamp, W., Huba, D., Ingerman, A., Ivanov, V., Kiddon, C., Konečný, J., Mazzocchi, S., McMahan, H. B., et al. (2019) · arXiv · doi:10.48550/arxiv.1902.01046 · Cited by 2,334 · cited 2×
“It has promising applications [14], [15] such as smart retail, smart healthcare, financial services, mobile content predictions, etc.”
BackgroundFederated Learning with Non-IID Data
Zhao, Y., Li, M., Lai, L., Suda, N., Civin, D., & Chandra, V. (2018) · arXiv · doi:10.48550/arxiv.1806.00582 · Cited by 2,077 · cited 1×
“, the data is generated by distinct distributions across different devices), imposes significant statistical challenges to fit a mode from the non-IID data [13], [17].”
Background
Research works printed but not linked: 2
- rockafellar, 2015, Convex Analysis
- tao, 1997, Convex analysis approach to DC programming: Theory, algorithms and applications, Acta Math Vietn, 22, 289
Cited by
Items 701–800 of 999
Zhou, R., Wang, Z., Yang, S., He, D., & Chan, S. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.114804 · Cited by 3
Liang, H., Wen, H., Wu, K., Letaief, K. B., & Xing, H. (2026) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2026.3740652 · Cited by 3
“Specifically, the problem is formulated as 4 ( P0 ) : min Given any { w ( t ) } T − 1 t =0 , the optimal s ( t ) i is given by [3], [35], s i , which yields Λ ( t ) = 0 , and φ t = 1 / (cid:107) w ( t ) (cid:107) 2 .”
InfluentialSun, X., Yuan, Z., Kong, X., Xue, L., Cheng, H., & Chen, Z. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3583081 · Cited by 4
“This strategy significantly reduces the communication rounds, thereby effectively reducing the overall communication overhead of FL. Yang et al. [33] proposed a novel FedAvg algorithm that calculates the global model by weighing each model according to the features of diverse devices.”
Optimal beamforming in over-the-air federated learning for efficient model aggregation
Choi, S., Kim, M., & Park, D. (2025) · ICT Express · doi:10.1016/j.icte.2025.06.016 · Cited by 3
Sha, Z., Huo, R., Sun, C., Wang, S., Huang, T., & Yu, F. (2025) · China Communications · doi:10.23919/jcc.fa.2023-0423.202507 · Cited by 3
“[67] proposes a fast model aggregation method using the superposition property of a wireless multiple-access channel.”
Towards Federated Learning Over the Air: Why Scaling Up Helps?
Zhu, J., Das, B., Παππάς, Ν., & Yang, H. H. (2025) · International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt) · doi:10.23919/wiopt66569.2025.11123201 · Cited by 3
“Moreover, an increase in the number of clients can improve energy efficiency, thereby reducing the transmit power [5]; in addition, analog over-the-air computing bypasses the encoding (resp. decoding) and modulation (resp. demodulation) processes, significantly reducing the access latency [3], [4].”
SegOTA: Accelerating Over-The-Air Federated Learning with Segmented Transmission
Zhang, C., Dong, M., Liang, B., Afana, A., & Ahmed, Y. (2025) · International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt) · doi:10.23919/wiopt66569.2025.11123377 · Cited by 2
“Various uplink beamforming designs have since been proposed to improve the training performance of wireless FL [9]–[13].”
Optimal Device Selection and Resource Allocation in Federated Learning
Kushwaha, D., & Hegde, R. M. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10888473 · Cited by 4
“However, the limited device and network resources restrict the communication frequency with the central server [3].”
Beamforming Design for Multi-Sector BD-RIS Assisted FL with AirComp
Xu, X., Ju, Y., Li, Z., Hou, X.-W., Liu, L., Mumtaz, S., & Wu, C. (2025) · IEEE Wireless Communications and Networking Conference · doi:10.1109/wcnc61545.2025.10978456 · Cited by 3
“In [7], the authors presented a method for joint device selection and receive beamforming design to enhance the communication efficiency and statistical performance of FL.”
Over-the-Air Federated Learning Under Time-Varying Wireless Channels Using OTFS
Rakesh, R. T. (2024) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2024.3465558 · Cited by 3
“A basic foundation of OTA computation based FL can be found in [4].”
LoLaFL: Low-Latency Federated Learning via Forward-Only Propagation
Zhang, J., Huang, J., & Huang, K. (2025) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2025.3625961 · Cited by 4
“Besides, the popular over-the-air computation (AirComp) technology is widely adopted to leverage the property of waveform superposition over a multi-access channel to realize simultaneous model uploading and over-the-air aggregation, thereby accelerating FL [24], [25], [26].”
Decentralized Over-the-Air Federated Learning in Full-Duplex MIMO Networks
Wang, Z., Zhou, Y., & Shi, Y. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3479769 · Cited by 4
“To obviate the inefficiency of orthogonal resource allocation in aggregating a large number of local models, over-the-air computation (Air-Comp) has recently gained notable prominence, given its capability of achieving low-latency model aggregation [8], [9].”
Self-citationJoint client selection and resource allocation for federated edge learning with imperfect CSI
Zhou, S., Wang, L., Wu, W., & Feng, L. (2024) · Computer Networks · doi:10.1016/j.comnet.2024.110914 · Cited by 4
Ultra-Reliable and Low-Latency Wireless Hierarchical Federated Learning: Performance Analysis
Zhang, H., Xu, P., & Dai, B. (2024) · Entropy · doi:10.3390/e26100827 · Cited by 3
“We assume that the channel matrix elements follow an i.i.d. distribution as CN ( 0, 1 ) [5,6,17].”
BackgroundMethodInfluentialChannel-Estimation-Free Gradient Aggregation for Over-the-Air SIMO Federated Learning
Zhong, C., & Yuan, X. (2024) · IEEE Wireless Communications Letters · doi:10.1109/lwc.2024.3382498 · Cited by 4
“Leveraging this property of FL, [1] proposed over-the-air FL (OA-FL), where the local gradients are aggregated in the gradient uploading via the wireless multiple access channel by exploiting the electromagnetic superposition, which sig-nificantly reduces the communication costs of the gradient…”
BackgroundWaveform Design for Over-the-Air Computing
Evgenidis, N. G., Mitsiou, N. A., Tegos, S. A., Diamantoulakis, P. D., Sarigiannidis, P. G., Rekanos, I. T., & Karagiannidis, G. K. (2025) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2025.3613838 · Cited by 4
“In this context, OTA computing has been studied as an effective way to provide the updates from the distributed devices to the FC while also achieving reduced convergence time [24], [25], [26], [27], [28].”
Power-Aware Sparse Reflect Beamforming in Active RIS-Aided Interference Channels
Long, R., Zhou, H., & Liang, Y. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3443173 · Cited by 4
“The replacement is valid when A is a positive semidefinite matrix [27].”
BackgroundInfluentialWireless Distributed Matrix-Vector Multiplication Using Over-the-Air Computation and Analog Coding
Choi, J. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3366547 · Cited by 3
“The server can receive and aggregate these transmitted vectors [9], [10], addressing the scalability issue associated with limited bandwidth.”
BackgroundOn-Cloud Linking Approach Using a Linkable Glue Layer for Metamorphic Edge Devices
Lee, D., & Park, D. (2023) · Electronics · doi:10.3390/electronics12244901 · Cited by 3
On the Privacy Leakage of Over-the-Air Federated Learning Over MIMO Fading Channels
Liu, H., Yan, J., & Zhang, Y.-J. A. (2023) · Global Communications Conference · doi:10.1109/globecom54140.2023.10437403 · Cited by 3
“In this work, we assume perfect downlink broadcasting owing to the adequate transmission power budget at the BS and focus on the uplink model aggregation as in [2], [4], [12], [14], [15].”
BackgroundMethodInfluentialQu, L., Song, S., Tsui, C.-Y., & Mao, Y. (2024) · IEEE Wireless Communications and Networking Conference · doi:10.1109/wcnc57260.2024.10570687 · Cited by 3
“While some previous research has tried to address this issue by focusing on the impact of communication error, it has been limited to analog communication [11]–[13].”
BackgroundTowards Scalable Wireless Federated Learning: Challenges and Solutions
Zhou, Y., Shi, Y.-M., Zhou, H., Wang, J., Fu, L., & Yang, Y. (2023) · arXiv.org · doi:10.48550/arxiv.2310.05076 · Cited by 3
“Hence, the devices with poor channel conditions, referred to as communication stragglers, lower the magnitude to be aligned at the edge server due to limited transmit power, thereby degrading the aggregation accuracy [8], [9].”
BackgroundSelf-citationA Survey of Edge Computing Resource Allocation Strategies Based on Federated Learning
Ni, S., He, Y., Chen, L., Wang, Y., & Yu, F. R. (2023) · International Conference on Networking and Network Applications (NaNA) · doi:10.1109/nana60121.2023.00027 · Cited by 6
“Yang uses Deep reinforcement learning (DQN) to optimize the mobility sensitive Federated learning network, so that the model owner can formulate the optimal energy and communication resource allocation scheme in a random and dynamic network environment [18].”
MethodDifferentially Private Wireless Federated Learning Using Orthogonal Sequences
Wei, X., Wang, T., Huang, R., Shen, C., Yang, J., & Poor, H. V. (2026) · IEEE Transactions on Information Theory · doi:10.1109/tit.2026.3668977 · Cited by 3
“The AirComp approach [7], [8], [9], [10] exploits the inherent signal superposition characteristics of a wireless multiple access channel to e ffi ciently perform sum / average computations.”
Over-the-Air Aggregation-Based Federated Learning in Cache-Enabled Wireless Edge Networks
Hong, J., Park, S., & Choi, W. (2023) · International Conference on Communications · doi:10.1109/icc45041.2023.10279408 · Cited by 5
“For the practical amplitude alignment in general channel model, a truncated channel inversion (TCI)-based power control method has been widely considered [5]–[7].”
MethodSari, N. I. P., Courtney, G., Torsten, P., & Jaroslaw, H. (2023) · Journal Markcount Finance · doi:10.55849/jmf.v1i1.56 · Cited by 4
Joint Receiver Design and User Scheduling for Over-the-Air Aggregation in Federated Learning
Zhang, F., Wan, J., Wang, K., & Wu, Q. (2023) · International Conference on Communications, Information System and Computer Engineering (CISCE) · doi:10.1109/cisce58541.2023.10142829 · Cited by 3
“Authors in [9] proposed a collaborative approach for efficient model aggregation using the over-the-air computation (AirComp) technique, which greatly improves the communication and computing efficiency.”
MethodOn-the-Fly Communication-and-Computing for Distributed Tensor Decomposition Over MIMO Channels
Chen, X., Larsson, E. G., & Huang, K. (2023) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2023.3329974 · Cited by 3
“AirComp finds applications in communication-efficient distributed computing and learning, and has been especially popular for federated learning (see, e.g., [15], [16], [17]).”
BackgroundOptimized Device Selection and Power Control for Wireless Federated Learning
Guo, W. M., Li, R., Huang, C. Z., Qin, X., Shen, K., & Zhang, W. (2022) · Globecom · doi:10.1109/globecom48099.2022.10001306 · Cited by 3
“Under the assumption of perfect downlink transmissions and only considering the uplink transmissions, existing works [5]–[7] investigated the device scheduling and power control design for AirCompbased FL by maximizing the number of selected devices under the mean square error (MSE) requirement of the aggregated signa…”
BackgroundMethodDifferential Optimization Federated Incremental Learning Algorithm Based on Blockchain
Chen, X., Luo, C., Wei, W., Xu, J., & Zhang, S. (2022) · Electronics · doi:10.3390/electronics11223814 · Cited by 3
“[5] proposed a vertical federated logistic regression method under the decentralized federated learning framework.”
MethodOver-the-air Clustered Wireless Federated Learning
Madhan-Sohini, A., Dominic, D., Shah, N., & Prasad, R. (2022) · 2023 IEEE Globecom Workshops (GC Wkshps) · doi:10.1109/gcwkshps58843.2023.10464705 · Cited by 3
Sarfraz, M., Alshahrani, H. M., Tarmissi, K. S., Alshahrani, H., Elfaki, M. A., Hamza, M. A., Nauman, A., & Khurshaid, T. (2022) · Sensors · doi:10.3390/s22228719 · Cited by 5
“[37] developed a federated learning algorithm that can be executed over-the-air.”
BackgroundOver-the-Air Gaussian Process Regression Based on Product of Experts
Sato, K. (2022) · 2022 IEEE Globecom Workshops (GC Wkshps) · doi:10.1109/gcwkshps56602.2022.10008771 · Cited by 3
“Over-the-air computation is a technique for communicationefficient distributed computation over shared channels based on nomographic functions [12], [13].”
MethodReconfigurable Intelligent Surface Assisted Over-the-Air Computation in Multi-Cell Networks
Li, J., Wu, Y., & Shi, Y.-M. (2022) · International Mediterranean Conference on Communications and Networking · doi:10.1109/meditcom55741.2022.9928626 · Cited by 3
“So far, AirComp has been applied in a wide range of applications related to distributed sensing [5], distributed edge learning[6], [7], [8], consensus control [9].”
Self-citationChen, Z., Tong, J., Fu, L., & Han, Z. (2022) · International Conference on Communications · doi:10.1109/icc45855.2022.9839106 · Cited by 2
“Without loss of generality, we adopt the proposition in [12] to design the transmitting beamformer, i.”
BackgroundMethodUnequal error protection transmission for federated learning
Zheng, S., & Chen, X. (2022) · IET Communications · doi:10.1049/cmu2.12379 · Cited by 3
Solving the Federated Edge Learning Participation Dilemma: A Truthful and Correlated Perspective
Hu, Q., Li, F., Zou, X., & Xiao, Y. (2022) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2022.3161099 · Cited by 3
Federated AirNet: Hybrid Digital-Analog Neural Network Transmission for Federated Learning
Fujihashi, T., Koike-Akino, T., & Watanabe, T. (2022) · arXiv.org · doi:10.48550/arxiv.2201.04557 · Cited by 3
“The analog over-the-air computation (AirComp) [8, 9] is another solution to adopt the analog modulation for the model parameter transmissions in FL.”
Enterprise Electricity Consumption Forecasting Method Based on Federated Learning
Zhai, Q., Zhang, X., & Cheng, J. (2021) · Communications in computer and information science · doi:10.1007/978-3-030-78618-2_46 · Cited by 4
Privacy-Preserving Wireless Federated Learning Exploiting Inherent Hardware Impairments
Aghdam, S. R., Amid, E., Furdek, M., & Amat, A. G. I. (2021) · IEEE International Workshop on Computer Aided Modeling and Design of Communication Links and Networks (CAMAD) · doi:10.1109/camad52502.2021.9617764 · Cited by 4
“This phenomenon is referred to as over-the-air aggregation, where simultaneously transmitted analog waves from different edge devices are weighted by channel coefficients and superposed at the server [3].”
BackgroundBehera, M. R., upadhyay, S., Otter, R., & Shetty, S. (2020) · doi:10.36227/techrxiv.13318163 · Cited by 2
“Weight of the participant plays an important role in federated learning environment [24] [25].”
BackgroundBlind Deconvolution Meets Phase Retrieval in Optical Wireless Communications
Fu, M., & Shi, Y. (2019) · IEEE Vehicular Technology Conference · doi:10.1109/vtcfall.2019.8891314 · Cited by 3
“Specifically, the exact DC representation for the rank-one constraint is that the difference between nuclear norm and Ky Fan k-norm [11] is equal to zero by setting k of the Ky Fan k-norm as one [12, 13].”
MethodSelf-citationJoint Client Scheduling and Resource Allocation Under Channel Uncertainty in Federated Learning
Wadu, M. M., Samarakoon, S., & Bennis, M. (2021) · IEEE Transactions on Communications · doi:10.1109/tcomm.2021.3088528 · Cited by 2
“The goal of FL is to train a highquality centralized model in a decentralized manner, based on local model training and client-server communication while training data remains private [3], [4], [6].”
BackgroundInfluentialAirPASS: Over-the-Air Federated Learning via Pinching Antenna Systems
Azimi-Abarghouyi, S. M., & Brinton, C. G. (2026) · arXiv.org · doi:10.48550/arxiv.2607.06768 · Cited by 2
“Substituting the above expression into the transmit power constraint yields [6] To satisfy all power constraints simultaneously, the denoising factor η must satisfy [6] η ≤ min Choosing η as the maximum feasible value minimizes the aggregation MSE.”
InfluentialA Unified Framework for Unbiased Non-Coherent Over-the-Air Computation
Dahl, M., Chen, Z., & Larsson, E. G. (2026) · IEEE Transactions on Communications · doi:10.1109/tcomm.2026.3698903 · Cited by 1
“Finally, θ t +1 is broadcast to all devices, and the FL loop continues [4].”
Analog RF Computing: A New Paradigm for Energy-Efficient Edge AI Over MU-MIMO SystemsPreprint
Yu, W., & Wong, V. W. S. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2605.14331 · Cited by 3
Azimi-Abarghouyi, S. M., Bennis, M., & Tassiulas, L. (2026) · arXiv.org · doi:10.48550/arxiv.2605.00931 · Cited by 2
“Orthogonal access can then create large bandwidth and latency overhead, which is why analog over-the-air aggregation and related simultaneous transmission strategies have become attractive in FL and HFL [8], [10], [11].”
Type-Based Unsourced Federated Learning With Client Self-Selection
Okumus, K., Ngo, K., Ganesan, U. K., Durisi, G., Ström, E. G., & Pandey, S. R. (2026) · International Conference on Communications · doi:10.1109/icc59461.2026.11587510 · Cited by 2
Age-Aware Edge-Blind Federated Learning via Over-the-Air Aggregation
Elshazly, A. E., & Arafa, A. M. (2026) · International Conference on Communications · doi:10.1109/icc59461.2026.11587676 · Cited by 1
“Recently, analog over-the-air (OTA) aggregation has been explored, exploiting the wireless multiple access channel (MAC)’s superposition property to allow simultaneous update transmissions that directly aggregate at the PS [20]–[23].”
Multi-UAV Deployment and Resource Allocation: A federated deep reinforcement learning approach
Zhu, Y., Ma, Y., Liu, X., Wang, H., & Yu, W. (2026) · IEEE Systems Man and Cybernetics Magazine · doi:10.1109/msmc.2025.3567942 · Cited by 3
“Furthermore, [22] achieved rapid global model aggregation by exploring the superposition characteristics of wireless multiple access channels.”
On Signal Peak Power Constraint of Over-the-Air Federated Learning
Bielefeld, L., Zheng, P., Hanay, O., Zhu, Y., Hu, Y.-L., & Schmeink, A. (2025) · doi:10.48550/arxiv.2512.23381 · Cited by 2
“We assume perfect CSI at the BS as commonly assumed in the literature [8], [6].”
Learned Digital Codes for Over-the-Air Computation in Federated Edge LearningPreprint
Tarizzo, A., Kazemi, M., & Gündüz, D. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2512.19777 · Cited by 2
“In [17], OTA computation is subsequently applied to FEEL with a multi-antenna server, jointly optimising device selection and receive beamforming to satisfy a mean-squared error (MSE) constraint while maximising the number of participating devices.”
Secure Over-the-Air Computation Against Multiple Eavesdroppers Using Correlated Artificial Noise
Nordlund, D., Maßny, L., Wachter-Zeh, A., Larsson, E. G., & Chen, Z. (2026) · IEEE Transactions on Communications · doi:10.1109/tcomm.2026.3742861 · Cited by 2
“Among its many promising applications in distributed systems, OtA FL has been extensively studied in recent works [11], [12], [13], [14], [15].”
Secure AirComp Federated Learning With STAR–IRS
Abyaneh, H. Z., Rabiei, A. M., & Ardakani, M. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3609447 · Cited by 2
“Similar to [10], [14], and [34], we assume a block fading channel model in which channel coe ffi cients are constant within a block but can vary independently among di ff erent blocks.”
Coherence-Aware Distributed Learning under Heterogeneous Downlink Impairments
Karbalayghareh, M., Love, D. J., & Brinton, C. G. (2026) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom59046.2026.11571334 · Cited by 2
“While digital FL emphasizes efficient scheduling and bandwidth allocation [15]–[19], OTA FL focuses on power control mechanisms to mitigate aggregation noise [20]–[26].”
Zema, P., Suraci, C., Molinaro, A., & Araniti, G. (2025) · IEEE International Conference on e-Health Networking, Applications and Services (Healthcom) · doi:10.1109/healthcom60686.2025.11342944 · Cited by 2
“Clients of the FL system are not required to exchange raw data, which ensures scalability and data privacy in large-scale deployments [4].”
Communication-Efficient Multi-Server Federated Learning via Over-the-Air Computation
Han, R., Ma, J., Bai, L., Choi, J., & Zhang, W. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3573600 · Cited by 3
“As proposed in [19], it is suggested to increase the number of users transmitting local gradient vectors.”
InfluentialRasheed, I., & Mostafa, H. (2025) · Computer Networks · doi:10.1016/j.comnet.2025.111553 · Cited by 3
Evolving Gaussian Systems as a Framework for Federated Regression Problems
Ožbot, M., Souza, P., & Škrjanc, I. (2025) · IEEE transactions on fuzzy systems · doi:10.1109/tfuzz.2025.3601900 · Cited by 2
“Applications of FL span a variety of critical sectors, including healthcare [6], geospatial services [7], [8], edge computing [2], [9], smart manufacturing [10], [11], and finance [12].”
RIS-Assisted XL-MIMO for Near-Field and Far-Field Communications
Cao, X., Mohammadi, M., Ngo, H. Q., Shin, H., & Matthaiou, M. (2025) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2025.3617656 · Cited by 3
“6 To this end, we find the lower bound for [37].”
Rethinking Federated Learning Over the Air: The Blessing of Scaling Up
Zhu, J., Das, B., Xie, Y. C., Παππάς, Ν., & Yang, H. H. (2025) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2025.3631503 · Cited by 1
“In addition, analog over-the-air computing bypasses the encoding (resp. decoding) and modulation (resp. demodulation) processes typically required before aggregation, which significantly reduces the access latency [5], [6], [10].”
Optimal Transceiver Design in Over-the-Air Federated Distillation
Hu, Z., Yan, J., Zhang, Y. A., Zhang, J., & Letaief, K. B. (2026) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2026.3742944 · Cited by 4
“The degradation of the learning performance caused by such inherent signal misalignment in the conventional over-the-air FL is dominated by the straggler with the worst channel condition and power capacity [11], [26].”
Personalized generative AI services through federated learning in 6G edge networks
Li, Z.-S., Chen, Z.-H., Hu, X.-Y., & Yang, H. (2025) · China Communications · doi:10.23919/jcc.fa.2024-0685.202507 · Cited by 2
“This bottleneck hampers training efficiency, restricting the scalability and performance of FL systems [32–33].”
FedABC: Attention-Based Client Selection for Federated Learning with Long-Term View
Ye, W., An, X., Wang, J., Yan, X., & Carle, G. (2025) · International Conference on Communications · doi:10.1109/icc52391.2025.11161564 · Cited by 1
“Prior work [15] has demonstrated that selecting more clients per global round can enhance FL performance, while resource constraints often hinder complete client participation.”
Air-FedGA: A Grouping Asynchronous Federated Learning Mechanism Exploiting Over-The-Air Computation
Ma, Q., Zhou, J., Hou, X., Liu, J., Xu, H., Miao, J., & Jia, Q. (2025) · IEEE International Parallel and Distributed Processing Symposium (IPDPS) · doi:10.1109/ipdps64566.2025.00051 · Cited by 3
“Yang et al. [16] focus on the trade-off between communication and learning, and propose a method that maximizes the number of devices while minimizing the mean squared error (MSE) of gradient error.”
Stealth Signals: Multi-Discriminator GANs for Covert Communications Against Diverse Wardens
Ali, A., Piran, M., & Arslan, H. (2025) · IEEE Transactions on Communications · doi:10.1109/tcomm.2026.3698908 · Cited by 2
Federated Learning-Driven Edge Intelligence Framework for Maritime Monitoring
Wu, J., Qiu, H., Liu, P., & Li, N. (2025) · International Conference on Sensor-Cloud and Edge Computing System (SCECS) · doi:10.1109/scecs65243.2025.11065527 · Cited by 2
“Finally for the optimization problem in the model aggregation process of cloud edge architecture, Yang et al. proposed an innovative federated learning model aggregation mechanism [7].”
Robust Over-The-Air Federated Learning In Heterogeneous Networks
Shaban, Z., Shah, N., & Prasad, R. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10890519 · Cited by 2
“Over-The-Air Federated Learning (OTA-FL) offers a solu-tion by using Over-the-Air Computation (AirComp), where the wireless channel acts as a natural aggregator, enabling simultaneous, non-orthogonal client transmissions [5], [6].”
Age-Aware Partial Gradient Update Strategy for Federated Learning Over the Air
Du, R., Zhu, J., Li, Z., & Yang, H. H. (2026) · International Conference on Communications · doi:10.1109/icc59461.2026.11587476 · Cited by 2
“To mitigate this issue, recent studies [2]–[5] have incorporated over-the-air (OTA) computation [6] into FL.”
RAG-based User Profiling for Precision Planning in Mixed-precision Over-the-Air Federated Learning
Yuan, J.-S., Tang, Y., & Guo, W.-S. (2025) · 2025 IEEE 102nd Vehicular Technology Conference (VTC2025-Fall) · doi:10.1109/vtc2025-fall65116.2025.11310436 · Cited by 2
“Over-the-Air Federated Learning (OTA-FL) [1] represents a novel paradigm in FL dedicated to wireless networks, which leverages the inherent randomness of physical-layer channel states and electromagnetic superposition for aggregating model updates.”
Improving Wireless Federated Learning via Joint Downlink-Uplink Beamforming Over Analog Transmission
Zhang, C., Dong, M., Liang, B., Afana, A., & Ahmed, Y. (2025) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2025.3595700 · Cited by 3
“Subsequently, various uplink beamforming designs have been proposed for analog schemes to improve the training performance of wireless FL [20]–[25].”
Sato, K., & Ishibashi, K. (2025) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2025.3562345 · Cited by 3
“The work in [29] proposed a joint device selection and beamforming design for AirComp-supported FL in wireless multiple-access channels.”
Pal, M. N., & De, D. (2025) · IET Quantum Communication · doi:10.1049/qtc2.70010 · Cited by 3
GP-FL: Model-Based Hessian Estimation for Second-Order Over-the-Air Federated LearningPreprint
Hamidi, S. M., Bereyhi, A., Asaad, S., & Poor, H. V. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2412.03867 · Cited by 2
“A well-known case is wireless FL, in which clients can use the superposition property of multiple access channel (MAC) to perform the computation directly over the air and significantly reduce the communication costs [12].”
Communication-Efficient Federated Learning over Wireless Channels via Gradient SketchingPreprint
Gattani, V. S., Zhang, J., & Dasarathy, G. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2410.23424 · Cited by 3
Liang, Y., Chen, Q., & Jiang, H. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2409.11240 · Cited by 2
“To enhance communication efficiency, over-the-air federated learning (OTA-FL) has emerged as a promising solution [13], [14].”
BackgroundTang, B., Hu, B., Qu, Z., & Ye, B. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3406918 · Cited by 2
Message Passing Based Wireless Federated Learning via Analog Message Aggregation
Zhang, Y., Xu, W., Liu, A., & Lau, V. K. N. (2024) · IEEE/CIC International Conference on Communications in China (ICCC) · doi:10.1109/iccc62479.2024.10682004 · Cited by 2
“Both analog model and gradient aggregations have been investigated for various application scenarios [5], [6].”
BackgroundXu, C., Wu, H., & Huang, K. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2407.20710 · Cited by 3
“Extensive research has applied AirComp in communication-efficient distributed computing and learning, such as AirComp-based federated learning in which a rich set of AirComp techniques are proposed to optimize the training efficiency [23]–[25].”
MethodScalable and Reliable Over-the-Air Federated Edge Learning
Egger, M., Hofmeister, C., Kaya, C., Bitar, R., & Wachter-Zeh, A. (2024) · Globecom · doi:10.1109/globecom52923.2024.10901264 · Cited by 3
Provision for Energy: A Resource Allocation Problem in Federated Learning for Edge Systems
Liu, M., Rajamanickam, L., & Parthasarathy, R. (2024) · EAI Endorsed Transactions on Energy Web · doi:10.4108/ew.6503 · Cited by 2
“In [28], a novel sparse and low-rank model was developed to improve statistical learning performance for on-device distributed training, and in [29], an energy-efficient bandwidth allocation scheme was proposed with constraints on learning performance.”
BackgroundBian, C., & Huang, G. (2024) · Environmental Monitoring & Assessment · doi:10.1007/s10661-024-12809-6 · Cited by 2
“…to traditional algorithms, federated learning approaches such as federated linear models, federated tree models, and federated support vector machines offer notable advantages in data transmission efficiency and model convergence (Huang & Ding, 2021; Liu et al., 2022; Yang et al., 2020; Yang 2023).”
Random Orthogonalization for Federated Learning in Massive MIMO Systems
Wei, X., Shen, C., Yang, J., & Poor, H. V. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3302335 · Cited by 22
“Restrictions apply. optimize FL has been studied in [10] and [11].”
BackgroundMethodInfluentialA Comparative Analysis of Federated and Centralized Machine Learning for Intrusion Detection in IoT
Khan, M. A., Rais, R. N. B., Khalid, O., & Khan, F. G. (2023) · International Arab Conference on Information Technology (ACIT) · doi:10.1109/acit58888.2023.10453751 · Cited by 3
“While recent studies have made strides in optimizing FL for bandwidth and communication efficiency [10][17], their assessments have primarily concentrated on privacy-sensitive scenarios.”
BackgroundOn Differential Privacy for Wireless Federated Learning with Non-coherent Aggregation
Seif, M., Şahin, A., Poor, H. V., & Goldsmith, A. J. (2023) · Globecom · doi:10.1109/globecom54140.2023.10437931 · Cited by 4
“To eliminate the need for CSI at the users, in [2] and [8], blind users are considered, and an aggregated CSI is used at the PS along with multiple antennas for computation.”
MethodEfficient Model Compression via Global Sparsification for Over-the-Air Federated Learning
Zheng, S., Dong, Y., & Chen, X. (2023) · IEEE Globecom Workshops (GC Wkshps) · doi:10.1109/gcwkshps58843.2023.10464512 · Cited by 3
“Another direction of recent works to improve the communication efficiency of FL is over-the-air computation (OAC) [5]–[7], which exploits the waveform superposition property of the multiple access channels (MAC) to perform model aggregating during transmission.”
BackgroundProbabilistic Device Scheduling for Over-the-Air Federated Learning
Sun, Y., Lin, Z., Mao, Y., Jin, S., & Zhang, J. (2023) · IEEE International Conference on Communication Technology (ICCT) · doi:10.1109/icct59356.2023.10419829 · Cited by 2
“To enable signal alignment in OTA computation, the selected edge devices in S t need to normalize their local gradients before transmission [2].”
MethodInfluentialFederated Linear Bandit Learning via Over-the-air Computation
Wang, J., Jiang, Y., Liu, X., Wang, T., & Shi, Y. (2023) · Globecom · doi:10.1109/globecom54140.2023.10437441 · Cited by 2
“Furthermore, the work in [14] states that AirComp realizes fast and spectrum-efficient model aggregation for wireless FL and thus decreases communication overheads.”
BackgroundMethodInfluentialSelf-citationInter-Cell Interference Coordination for AirComp Federated Learning in Cellular Networks
Zhang, R.-Q., Deng, N., & Wei, H.-C. (2023) · 2023 IEEE/CIC International Conference on Communications in China (ICCC) · doi:10.1109/iccc57788.2023.10233371 · Cited by 2
Virtual Experience-Based Mobile Device Selection Algorithm for Federated Learning
Paik, M., Ko, H., & Pack, S. (2022) · IEEE Systems Journal · doi:10.1109/jsyst.2022.3206404 · Cited by 2
“[12] conducted a joint device selection and beamforming design to increase the number of selected MDs while satisfying the mean-square-error (MSE) requirement.”
BackgroundHybrid Reconfigurable Intelligent Surface Assisted Over-the-Air Federated Learning
Jin, J., Wang, Z., Wu, L., & Zhou, Y. (2023) · IEEE International Conference on Communications workshops · doi:10.1109/iccworkshops57953.2023.10283597 · Cited by 2
“Specifically, the signals from multiple distributed edge devices are aggregated via AirComp through the natural superposition property of analog signals in multiple-access channels [7]–[9].”
BackgroundMethodAdaptive federated learning aggregation strategies based on mobile edge computing
Liu, S., & Xu, F. (2023) · Conference on Machine Learning and Computer Application · doi:10.1117/12.2675238 · Cited by 2
“…of classical machine learning models in the federated learning framework, such as the federated logistic regression proposed by Yang et al [6], the federated forest proposed by Lin et al [7] and the federated SVM proposed by Hartmann et al [8], and similarly, there are scholars have…”
MethodInerrant-Bits Consumption for Over-the-Air Computation-based Multi-tier Computing
Wan, J., Zhang, F., & Wang, K. (2023) · International Conference on Communications, Information System and Computer Engineering (CISCE) · doi:10.1109/cisce58541.2023.10142476 · Cited by 2
“The authors in [7] proposed a new AirComp-based fast global model aggregation for Federated Learning (FL) systems, which is a distributed machine learning technique.”
BackgroundBreaking Hardware Boundaries of IoT Devices via Inverse Feature Completion
Gao, Y., Ong, Y.-S., Gong, M., Jiang, F., Zhang, Y., & Wang, S. (2022) · IEEE Internet of Things Journal · doi:10.1109/jiot.2022.3222191 · Cited by 2
“There has been a growing interest in the privacy-preserving collaborative learning frameworks [5], [6] in the past few years.”
BackgroundEnergy efficiency of full‐duplex communication system assisted by reconfigurable intelligent surface
Wang, Y., Guan, P., Yu, H., & Zhao, Y. (2022) · IET Communications · doi:10.1049/cmu2.12558 · Cited by 2
Jamming-Resilient Over-the-Air Computation for Federated Learning in Edge Intelligence
Zou, Y., Xu, M., Li, H., Hu, Q., Deng, X., Yu, D., & Cheng, X. (2022) · IEEE Smart World Congress (SWC) · doi:10.1109/smartworld-uic-atc-scalcom-digitaltwin-pricomp-metaverse56740.2022.00038 · Cited by 2
Federated Learning with Blockchain: A Study of the Latest Decentralized Couple
Faridi, A. R., Hafeez, A., & Masood, F. (2022) · International Conference on Computing, Communication, and Intelligent Systems (ICCCIS) · doi:10.1109/icccis56430.2022.10037647 · Cited by 3
Su, L., & Lau, V. K. N. (2022) · IEEE Internet of Things Journal · doi:10.1109/jiot.2022.3177449 · Cited by 4
DFHelper: Help clients to participate in federated learning tasks
Wu, Z., Gao, J., Zhang, J., Li, Y., Li, Q., Guan, Z., & Zhong, C. (2022) · Applied Intelligence · doi:10.1007/s10489-022-04081-3 · Cited by 2
Multi-Task Federated Edge Learning (MTFeeL) With SignSGD
Mahara, S. S., Shruti, M., & Bharath, B. N. (2022) · National Conference on Communications · doi:10.1109/ncc55593.2022.9806778 · Cited by 5
“In wireless edge devices, the number of updates to each device can be further reduced by exploiting the nature of the wireless medium leading to a solution called over-the-air aggregation also called over-theair computation [11]–[13].”
Background
Metadata
- Authors
- Yang, K., Jiang, T., Shi, Y., & Ding, Z.
- Affiliations
- Chinese Academy of Sciences (CN)
- ShanghaiTech University (CN)
- University of California, Davis (US)
- Venue
- IEEE Transactions on Wireless Communications
- Published
- 2020-03
- Volume
- 19
- Issue
- 3
- Pages
- 2022-2035
- ISSN
- 1536-1276 (ISSN-L)
- 1558-2248
- OpenAlex
- W2999074226
- Cited by
- 1,084
- References
- 62
- Citations by year
- 2026: 86 · 2025: 154 · 2024: 176 · 2023: 190 · 2022: 189 · 2021: 190 · 2020: 80
- FWCI
- 63.79
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE Transactions on Wireless Communications (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- National Natural Science Foundation of China (61601290)
- National Science Foundation (CNS-1702752, ECCS-1711823)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data
- Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Advanced Wireless Communication Technologies
- Keywords
- federated learning, over-the-air computation, edge machine learning, model aggregation, device selection, beamforming design, difference of convex functions, DC algorithm, non-IID data, data privacy
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.