Optimal Wireless Resource Allocation With Random Edge Graph Neural Networks
Open accessThe DOI matches its Crossref record; 40 of the 41 references it cites were found in a registry or an index.
Abstract
We consider the problem of optimally allocating resources across a set of transmitters and receivers in a wireless network. The resulting optimization problem takes the form of constrained statistical learning, in which solutions can be found in a model-free manner by parameterizing the resource allocation policy. Convolutional neural networks architectures are an attractive option for parameterization, as their dimensionality is small and does not scale with network size. We introduce the random edge graph neural network (REGNN), which performs convolutions over random graphs formed by the fading interference patterns in the wireless network. The REGNN-based allocation policies are shown to retain an important permutation equivariance property that makes them amenable to transference to different networks. We further present an unsupervised model-free primal-dual learning algorithm to train the weights of the REGNN. Through numerical simulations, we demonstrate the strong performance REGNNs obtain relative to heuristic benchmarks and their transference capabilities.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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
“Graph neural network (GNN) has recently been shown to be able to harness the benefits of generalizability, interpretability, robustness, scalability, superior performance, real-time and distributed implementation for learning to optimize nonconvex problems, including power control [270], beamforming [23], and phase shift design [25].”
From Edge artificial intelligence for 6g: vision, enabling technologies, and applications (2022) · cites this paper as [270]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 2×
“In fast fading scenarios the instantaneous value of the reward r(t) in (1) is not directly experienced by end users.”
BackgroundBertsekas, D. P. (1997) · Journal of the Operational Research Society · doi:10.1038/sj.jors.2600425 · Cited by 11,726
(1999) · Cited by 5,147
Wide & Deep Learning for Recommender Systems
Cheng, H.-T., Koc, L., Harmsen, J., Shaked, T., Chandra, T., Aradhye, H., Anderson, G., Corrado, G., Chai, W., Ispir, M., et al. (2016) · Proceedings of the 1st Workshop on Deep Learning for Recommender Systems · doi:10.1145/2988450.2988454 · Cited by 3,535
Shi, Q., Razaviyayn, M., Luo, Z.-Q., & He, C. (2011) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2011.2147784 · Cited by 2,129
Dual methods for nonconvex spectrum optimization of multicarrier systems
Yu, W., & Lui, R. (2006) · IEEE Transactions on Communications · doi:10.1109/tcomm.2006.877962 · Cited by 1,572 · cited 1×
“Some simplification is attained by working in the Lagrangian dual domain [1], [2] and subsequently using dual descent methods [3]–[5], or, alternatively, with heuristic optimization and scheduling methods [6]–[9].”
MethodDeep Convolutional Networks on Graph-Structured Data
Henaff, M., Bruna, J., & LeCun, Y. (2015) · arXiv · doi:10.48550/arxiv.1506.05163 · Cited by 1,255
Learning to Optimize: Training Deep Neural Networks for Interference Management
Sun, H., Chen, X., Shi, Q., Hong, M., Fu, X., & Sidiropoulos, N. D. (2018) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2018.2866382 · Cited by 874
Shi, Q., Razaviyayn, M., Luo, Z.-Q., & He, C. (2011) · 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2011.5946304 · Cited by 873
Cited by
Edge artificial intelligence for 6g: vision, enabling technologies, and applications
Letaief, K. B., Shi, Y., Lu, J., & Lu, J. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2021.3126076 · Cited by 829
“…been shown to be able to harness the benefits of generalizability, interpretability, robustness, scalability, superior performance, real-time and distributed implementation for learning to optimize nonconvex problems, including power control [270], beamforming [23], and phase shift design [25].”
BackgroundShe, C., Sun, C., Gu, Z., Li, Y., Yang, C., Poor, H. V., & Vucetic, B. (2021) · Proceedings of the IEEE · doi:10.1109/jproc.2021.3053601 · Cited by 405
“The details on the model-free unsupervised deep learning method can be found in [63], [232], [233], and [242].”
BackgroundMethodShen, Y., Shi, Y., Zhang, J., & Letaief, K. B. (2021) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2020.3036965 · Cited by 367
“cannot be solved by existing methods in [10], [13], [14].”
BackgroundMethodInfluentialGraph-based deep learning for communication networks: A survey
Jiang, W. (2022) · Computer Communications · doi:10.1016/j.comcom.2021.12.015 · Cited by 284
“To handle this problem, multiple GNN-based solutions are proposed [83, 58, 40, 57, 42, 112, 82, 32, 84].”
BackgroundNetwork Digital Twin: Context, Enabling Technologies, and Opportunities
Almasan, P., Ferriol-Galmes, M., Paillisse, J., Suarez-Varela, J., Perino, D., Lopez, D., Perales, A. A. P., Harvey, P., Ciavaglia, L., Wong, L., et al. (2022) · IEEE Communications Magazine · doi:10.1109/mcom.001.2200012 · Cited by 180
“In this context, GNN models have recently shown promising results for generalization across network-related data structured as graphs [2], [6].”
BackgroundMethodInfluentialJiang, T., Cheng, H. V., & Yu, W. (2021) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2021.3078502 · Cited by 290
Deep-Learning-Based Wireless Resource Allocation With Application to Vehicular Networks
Liang, L., Ye, H., Yu, G., & Li, G. Y. (2020) · Proceedings of the IEEE · doi:10.1109/jproc.2019.2957798 · Cited by 282
“That said, we notice there have been some recent studies [91], [92] showing graph neural networks, or graph convolutional networks [93], can be particularly efficient in modeling wireless interference, which might represent a good start in this direction.”
ResultMazhar, T., Irfan, H. M., Haq, I., Ullah, I., Ashraf, M., Shloul, T. A., Ghadi, Y. Y., Imran, & Elkamchouchi, D. H. (2023) · Electronics · doi:10.3390/electronics12010242 · Cited by 237
Graph Neural Networks for Wireless Communications: From Theory to Practice
Shen, Y., Zhang, J., Song, S. H., & Letaief, K. B. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3219840 · Cited by 230
“Second, the qualitative results of [3], [12], [16], [20] cannot be used to guide a better design of GNN-based algorithms, i.”
BackgroundMethodInfluentialMachine Learning for Large-Scale Optimization in 6G Wireless Networks
Shi, Y., Lian, L., Shi, Y., Wang, Z., Zhou, Y., Fu, L., Bai, L., Zhang, J., & Zhang, W. (2023) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2023.3300664 · Cited by 205
“Power Allocation [85] BS / User Communication link State information Channel coefficient”
BackgroundMethodInfluentialAdversarial Machine Learning in Wireless Communications Using RF Data: A Review
Adesina, D., Hsieh, C.-C., Sagduyu, Y. E., & Qian, L. (2023) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2022.3205184 · Cited by 174
“Numerous DL-based algorithms are being developed to make resource allocation more efficient [11], [121], [122], [123], [124], [125].”
Optimal Power Flow Using Graph Neural Networks
Owerko, D., Gama, F., & Ribeiro, A. (2020) · ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp40776.2020.9053140 · Cited by 160
“In particular, we focus on graph neural networks (GNNs) [13–15] to exploit their stability properties [21] that guarantee scalability [22, 23].”
BackgroundSelf-citationA Survey of Recent Advances in Optimization Methods for Wireless Communications
Liu, Y., Chang, T., Hong, M., Wu, Z., So, A. M., Jorswieck, E. A., & Yu, W. (2024) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2024.3443759 · Cited by 225
“In [208], the fully connected networks used in the previous works are replaced by certain random edge graph neural network (REGNN), which performs convolu-tions on random graphs created by the network’s fading interference patterns.”
BackgroundMethodZhou, H., Erol‐Kantarci, M., Liu, Y., & Poor, H. V. (2023) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2023.3340099 · Cited by 210
“Therefore, graph learning has been applied to power control and interference management [163], resource allocation [164], [165], network slicing [166], and so on.”
BackgroundGraph Filters for Signal Processing and Machine Learning on Graphs
Isufi, E., Gama, F., Shuman, D. I., & Segarra, S. (2024) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2024.3349788 · Cited by 217
“…networks where nodes are buses and edges are power lines [34]; iii ) telecommunication networks where nodes are transceivers and edges are channels [35], [36]; iv ) water networks where nodes are junctions and edges are pipes [37], and; v ) road networks where nodes are intersections and edges are…”
Quy, V. K., Chehri, A., Quy, N. M., Han, N. D., & Ban, N. T. (2023) · IEEE Access · doi:10.1109/access.2023.3269297 · Cited by 164
“…CNNs (Convolutional Neural Networks) and GNNs (Graph Neural Networks), have been strongly applied in computing and information processing and achieved some breakthrough results in several areas such as recommendation systems [147], collaboration control systems [148], and wireless networks [149].”
BackgroundUnfolding WMMSE Using Graph Neural Networks for Efficient Power Allocation
Chowdhury, A., Verma, G., Rao, C., Swami, A., & Segarra, S. (2021) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2021.3071480 · Cited by 162
“It should be noted that, unlike [23] and in accordance with some recent works [25], [26], UWMMSE is an unsupervised method since it requires access to samples of the channel state matrices H but does not require access to the optimal power allocations (labels) associated with those channels.”
BackgroundMethodResultInfluentialThe Role of Machine Learning and the Internet of Things in Smart Buildings for Energy Efficiency
Shah, S. B., Iqbal, M., Aziz, Z., Rana, T. A., Khalid, A., Cheah, Y., & Arif, M. S. (2022) · Applied Sciences · doi:10.3390/app12157882 · Cited by 182
A GNN-Based Supervised Learning Framework for Resource Allocation in Wireless IoT Networks
Chen, T., Zhang, X., You, M., Zheng, G., & Lambotharan, S. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3091551 · Cited by 175
“Additionally, a random edge GNN was proposed in [24] to solve the power optimization problems in wireless ad-hoc networks and cellular networks.”
BackgroundMethodGraph Neural Networks: Architectures, Stability, and Transferability
Ruiz, L., Gama, F., & Ribeiro, A. (2021) · Proceedings of the IEEE · doi:10.1109/jproc.2021.3055400 · Cited by 216
“15 and 16 show training curves for the solution of (55) with an REGNN parameterization [58].”
BackgroundMethodInfluentialSelf-citationFault diagnosis of power transformers using graph convolutional network
Liao, W., Yang, D., Wang, Y., & Ren, X. (2020) · CSEE Journal of Power and Energy Systems · doi:10.17775/cseejpes.2020.04120 · Cited by 149
An Overview on the Application of Graph Neural Networks in Wireless Networks
He, S., Xiong, S., Ou, Y., Zhang, J., Wang, J., Huang, Y., & Zhang, Y. (2021) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2021.3128637 · Cited by 174
Graph Neural Networks for Communication Networks: Context, Use Cases and Opportunities
Suárez‐Varela, J., Almasan, P., Ferriol-Galmés, M., Rusek, K., Geyer, F., Cheng, X., Shi, X., Xiao, S., Scarselli, F., Cabellos‐Aparicio, A., et al. (2022) · IEEE Network · doi:10.1109/mnet.123.2100773 · Cited by 164
Learning Power Allocation for Multi-Cell-Multi-User Systems With Heterogeneous Graph Neural Networks
Guo, J., & Yang, C. (2021) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2021.3100133 · Cited by 148
“Recently, the PE and PI properties of wireless tasks have been noticed in [9], [10], [24].”
BackgroundMethodInfluentialAI Empowered Wireless Communications: From Bits to Semantics
Qin, Z., Liang, L., Wang, Z., Jin, S., Tao, X., Tong, W., & Li, G. Y. (2024) · Proceedings of the IEEE · doi:10.1109/jproc.2024.3437730 · Cited by 155
Embedding Graph Auto-Encoder for Graph Clustering
Zhang, H., Li, P., Zhang, R., & Li, X. (2022) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2022.3158654 · Cited by 135
Cooperative Trajectory Design of Multiple UAV Base Stations With Heterogeneous Graph Neural Networks
Zhang, X., Zhao, H., Wei, J., Yan, C., Xiong, J., & Liu, X. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3204794 · Cited by 129
“To enable more flexible and efficient representation of complex relations among UAV-BSs and GTs, we turn to GNN. Pioneer applications of GNNs in wireless communications are [32], [33], and [34], where wireless links and their interference relations are built into graphs.”
MethodGu, Z., She, C., Hardjawana, W., Lumb, S., McKechnie, D., Essery, T., & Vucetic, B. S. (2021) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2021.3078498 · Cited by 101
Liu, Z., Zhang, J., Shi, E., Liu, Z., Niyato, D. T., Ai, B., & Shen, X. (2024) · IEEE Wireless Communications · doi:10.1109/mwc.015.2300595 · Cited by 96
Pinching-Antenna Systems (PASS): A Tutorial
Liu, Y., Jiang, H., Xu, X., Wang, Z., Guo, J., Ouyang, C., Mu, X., Ding, Z., Nallanathan, A., Karagiannidis, G. K., et al. (2026) · IEEE Transactions on Communications · doi:10.1109/tcomm.2026.3658289 · Cited by 150
“It has been noticed that the advantages of GNNs stem from harnessing the permutation properties that widely exist in wireless policies [112]–[115].”
Learning Decentralized Wireless Resource Allocations With Graph Neural Networks
Wang, Z., Eisen, M., & Ribeiro, A. (2022) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2022.3163626 · Cited by 100
“With the propositions brought out above, we can state the theorem from [19] as follows.”
InfluentialSelf-citationAnomaly Detection in Cloud Computing using Knowledge Graph Embedding and Machine Learning Mechanisms
Mitropoulou, K., Kokkinos, P. C., Soumplis, P., & Varvarigos, E. (2023) · Journal of Grid Computing · doi:10.1007/s10723-023-09727-1 · Cited by 94
“…and ML techniques towards automated resource allocation is mostly unexplored, with a few notable exceptions that leverage edge graph neural networks for the construction of allocation policies [50] and the prediction of corresponding Quality of Service (QoS) performance indicators [51].”
Scalable Power Control/Beamforming in Heterogeneous Wireless Networks with Graph Neural Networks
Zhang, X., Zhao, H., Xiong, J., Liu, X., Zhou, L., & Wei, J. (2021) · Globecom · doi:10.1109/globecom46510.2021.9685457 · Cited by 90
“This work could be considered as a generalization of [9], [12] from homogeneous networks to more complex heterogeneous systems.”
BackgroundInfluentialA Bipartite Graph Neural Network Approach for Scalable Beamforming Optimization
Kim, J., Lee, H., Hong, S., & Park, S.-H. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3193138 · Cited by 84
“, transmitter-receiver pairs [20]–[22], receive antennas [23], and users [24].”
BackgroundInfluentialDistributed Scheduling Using Graph Neural Networks
Zhao, Z., Verma, G., Rao, C. R., Swami, A., & Segarra, S. (2021) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp39728.2021.9414098 · Cited by 89
“Here, we adopt an alternative direction [20, 24], where graph neural networks (GNNs) [25–28] are used to incorporate the topology of the wireless network in the learning algorithm.”
MethodGraphon Neural Networks and the Transferability of Graph Neural NetworksPreprint
Ruiz, L., Chamon, L. F. O., & Ribeiro, A. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2006.03548 · Cited by 164
Self-citationA Graph Neural Network Learning Approach to Optimize RIS-Assisted Federated Learning
Wang, Z., Zhou, Y., Zou, Y., An, Q., Shi, Y., & Bennis, M. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3239400 · Cited by 77
“By leveraging the permutation equivalence and invariance, the authors in [33] developed a random edge GNN for power allocation for wireless networks with frequency reuse, which was extended to the scenario with multiple antennas in [34].”
BackgroundMethodSun, R., Cheng, N., Li, C., Quan, W., Zhou, H., Wang, Y., Zhang, W., & Shen, X. (2025) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2025.3574765 · Cited by 89
“Therefore, this approach has been applied in probabilistic constrained power allocation for video streaming transmission [191], joint power and bandwidth allocation for ultra-reliable and low-latency communications [192], and power allocation with statistical constraints in ad-hoc networks [62].”
Link Scheduling Using Graph Neural Networks
Zhao, Z., Verma, G., Rao, C. R., Swami, A., & Segarra, S. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3222781 · Cited by 87
AI-Driven Collaborative Resource Allocation for Task Execution in 6G-Enabled Massive IoT
Lin, K., Li, Y., Zhang, Q., & Fortino, G. (2021) · IEEE Internet of Things Journal · doi:10.1109/jiot.2021.3051031 · Cited by 70
“Eisen and Ribeiro [2] designed a policy for resource allocation across a set of transmitters and receivers in a wireless network.”
BackgroundMachine Learning for Massive Industrial Internet of Things
Zhou, H., She, C., Deng, Y., Döhler, M., & Nallanathan, A. (2021) · IEEE Wireless Communications · doi:10.1109/mwc.301.2000478 · Cited by 66
Decentralized Inference With Graph Neural Networks in Wireless Communication Systems
Lee, M., Yu, G., & Dai, H. (2021) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2021.3125793 · Cited by 66
“The first category of works focus on using GNNs for different problems in wireless communication systems, such as link scheduling [10], power allocation [28], [29] and user association [30].”
BackgroundMultidimensional Graph Neural Networks for Wireless Communications
Liu, S., Guo, J., & Yang, C. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3305124 · Cited by 83
“A GNN with the same permutation property as a policy can learn the policy efficiently, and is possible to be scalable to large-scale problems and generalizable to unseen system scales [7]–[9].”
BackgroundMethodInfluentialTam, P., Song, I., Kang, S., Ros, S., & Kim, S. (2022) · Electronics · doi:10.3390/electronics11203371 · Cited by 66
“In [82], an optimal wireless resource allocation was investigated with the primary approach using random edge GNN.”
MethodInfluentialRuiz, L., Chamon, L. F. O., & Ribeiro, A. (2021) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2021.3106857 · Cited by 98
Self-citationGu, Y., She, C., Quan, Z., Qiu, C., & Xu, X. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3253126 · Cited by 63
“In [21], the authors considered a large-scale dynamic wireless network with a set of transmitters and receivers, and formulated a generic resource allocation problem.”
BackgroundLearning Resource Allocation Policy: Vertex-GNN or Edge-GNN?
Peng, Y., Guo, J., & Yang, C. (2024) · IEEE Transactions on Machine Learning in Communications and Networking · doi:10.1109/tmlcn.2024.3354872 · Cited by 59
“In the literature of intelligent communications, GNNs have been designed to learn the link scheduling [4], [5], [15], [16], power control [6], [17] and power allocation [9] policies in device-to-device (D2D) communications and interference networks, the precoding policies in multi-user…”
BackgroundWang, Y., Li, Y., Shi, Q., & Wu, Y. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3325735 · Cited by 62
Coordinated Sum-Rate Maximization in Multicell MU-MIMO With Deep Unrolling
Schynol, L., & Pesavento, M. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2023.3242716 · Cited by 58
“Unlike other approaches [15], [25], [27] based on GNNs, in which the graph underlying the GNN mimicks the wireless network topology, our GNN graph shift matrices represent transceiver signal spaces to achieve generality in the number of transceiver antennas.”
MethodJagannath, J., Ramezanpour, K., & Jagannath, A. (2022) · ACM Workshop on Wireless Security and Machine Learning (WiseML) · doi:10.1145/3522783.3529519 · Cited by 59
“The capability of GNN in modeling spatial information has been commonly used in statistical learning techniques for resource allocation in wireless networks [9].”
GNN-Based Power Allocation and User Association in Digital Twin Network for the Terahertz Band
Zhang, H., Ma, X., Liu, X., Li, L., & Sun, K. (2023) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2023.3313192 · Cited by 55
“Using this property, the policy trained in the small-scale network can be applied to the large-scale network without changing the performance [13].”
MethodA new 5G radio evolution towards 5G-Advanced
Pang, J., Wang, S., Tang, Z., Qin, Y., Tao, X., You, X., & Zhu, J. (2022) · Science China Information Sciences · doi:10.1007/s11432-021-3470-1 · Cited by 50
Graph Neural Networks Meet Wireless Communications: Motivation, Applications, and Future Directions
Lee, M., Yu, G., Dai, H., & Li, G. Y. (2022) · IEEE Wireless Communications · doi:10.1109/mwc.001.2200023 · Cited by 64
“On the one hand, as a powerful tool for wireless communications, GNNs have been applied to solve various tasks in wireless communications, including link scheduling [6], power control [7]–[9], user association [10], and other emerging fields [11], [12].”
MethodSun, H., Pu, W., Fu, X., Chang, T., & Hong, M. (2022) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2022.3143372 · Cited by 59
“Different network structures, such as convolutional neural networks [14] and graph neural networks [16, 34], and different modeling techniques, such as reinforcement learning [35], are also studied”
BackgroundYang, H., Cheng, N., Sun, R., Quan, W., Chai, R., Aldubaikhy, K., Alqasir, A., & Shen, X. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3368516 · Cited by 55
“A random edge GNN (REGNN) is proposed to enhance scalability and generalization for optimal power control in interference channels [21].”
BackgroundMethodLearning Resilient Radio Resource Management Policies With Graph Neural Networks
Naderializadeh, N., Eisen, M., & Ribeiro, A. (2023) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2023.3255547 · Cited by 54
“More recently, the authors in [15], [25] considered robust formulations of the RRM problem, where arbitrary constraints can be included in the optimization problem, such as per-user minimum-capacity requirements.”
BackgroundMethodSelf-citationToward Scalable Wireless Federated Learning: Challenges and Solutions
Zhou, Y., Shi, Y., Zhou, H., Wang, J., Fu, L., & Yang, Y. (2023) · IEEE Internet of Things Magazine · doi:10.1109/iotm.001.2300099 · Cited by 44
Li, B., Yang, L., Maunder, R. G., Sun, S., & Xiao, P. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3284263 · Cited by 51
“Additionally, the authors in [14] and [15] studied the GNN-assisted PA in ad hoc networks and the proposed methods significantly outperform the conventional methods owing to the GNN’s generalizability.”
A Survey of Graph-Based Resource Management in Wireless Networks—Part II: Learning Approaches
Dai, Y., Lyu, L., Cheng, N., Sheng, M., Liu, J., Wang, X., Cui, S. R., Cai, L., & Shen, X. (2024) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2024.3508777 · Cited by 55
“Thus, REGNN for this access control problem can inherit its design for power control [2], [71].”
Graph Neural Networks Approach for Joint Wireless Power Control and Spectrum Allocation
Marwani, M., & Kaddoum, G. (2024) · IEEE Transactions on Machine Learning in Communications and Networking · doi:10.1109/tmlcn.2024.3408723 · Cited by 46
“Similarly, [33] introduced a GNN-based framework for learning resource allocation strategies in wireless networks, offering reduced training times and improved scalability compared to conventional MLPs. Nonetheless, this framework was not well-suited for heterogeneous wireless devices or systems…”
BackgroundA Survey of Graph-Based Resource Management in Wireless Networks—Part I: Optimization Approaches
Dai, Y., Lyu, L., Cheng, N., Sheng, M., Liu, J., Wang, X., Cui, S. R., Cai, L., & Shen, X. (2024) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2024.3508783 · Cited by 44
“Eisen and Ribeiro first employed GNN in wireless networks to solve link scheduling problems in multi-hop networks and multiple access scheduling problems in cellular networks [13], [14].”
GBLinks: GNN-Based Beam Selection and Link Activation for Ultra-Dense D2D mmWave Networks
He, S., Xiong, S., Zhang, W., Yang, Y., Ren, J., & Huang, Y. (2022) · IEEE Transactions on Communications · doi:10.1109/tcomm.2022.3158646 · Cited by 46
Modular Meta-Learning for Power Control via Random Edge Graph Neural Networks
Nikoloska, I., & Simeone, O. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3195352 · Cited by 42
“• We validate the performance of all meta-learning methods with extensive experiments that provide comparisons with joint training schemes [9].”
BackgroundMethodInfluentialGraph Convolutional Value Decomposition in Multi-Agent Reinforcement LearningPreprint
Naderializadeh, N., Hung, F., Soleyman, S., & Khosla, D. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.04740 · Cited by 44
“Due to the flexibility in modelling structured data, GNNs have seen widespread application, in areas such as knowledge representation [19], natural language processing [20], social network analysis [21], wireless communications [22], chemistry [23], and physics [24].”
Unsupervised Optimal Power Flow Using Graph Neural Networks
Owerko, D., Gama, F., & Ribeiro, A. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10446827 · Cited by 44
“Meanwhile, we use the square function as the penalty for the equality constraints:”
MethodSelf-citationEdge Graph Intelligence: Reciprocally Empowering Edge Networks With Graph Intelligence
Zeng, L., Ye, S., Chen, X., Zhang, X., Ren, J., Tang, J., Yang, Y., & Shen, X. (2025) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2025.3527561 · Cited by 47
“Their approach is complemented by Eisen et al. [239], [240], who employed spectral-based GCNs for power allocation in Device-to-Device (D2D) wireless networks, highlighting the effectiveness of GCNs in managing interference.”
Guo, S., Han, K., Gong, W., Li, L., Tian, F., & Jiang, X. (2022) · Sensors · doi:10.3390/s22239304 · Cited by 35
Wireless Power Control via Counterfactual Optimization of Graph Neural Networks
Naderializadeh, N., Eisen, M., & Ribeiro, A. (2020) · IEEE International Workshop on Signal Processing Advances in Wireless Communications (SPAWC) · doi:10.1109/spawc48557.2020.9154336 · Cited by 35
Self-citationLearning with Limited Samples: Meta-Learning and Applications to Communication Systems
Chen, L., Jose, S. T., Nikoloska, I., Park, S., Chen, T., & Simeone, O. (2023) · Foundations and Trends® in Signal Processing · doi:10.1561/2000000115 · Cited by 46
Opportunities and challenges of graph neural networks in electrical engineering
Chien, E., Li, M., Aportela, A., Ding, K., Jia, S., Maji, S., Zhao, Z., Duarte, J. M., Fung, V., Hao, C., et al. (2024) · Nature Reviews Electrical Engineering · doi:10.1038/s44287-024-00076-z · Cited by 36
“Notable examples include resource allocation problems in wireless networks 10,12,13,65 and optimal power flow in power grids 16,18,66 , as well as the low-density parity-check decoding problem where the relationship between variable nodes and check nodes form a graph 67,68 .”
Deep Graph Unfolding for Beamforming in MU-MIMO Interference Networks
Chowdhury, A., Verma, G., Swami, A., & Segarra, S. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3323207 · Cited by 41
“Several neural architectures of varying complexity – multi-layer perceptrons (MLP) [15], [16], convolutional neural networks (CNN) [17], recurrent neural networks (RNN) [18], and even graph neural networks (GNN) [19], [20] – have been applied to this end.”
State-Augmented Learnable Algorithms for Resource Management in Wireless Networks
Naderializadeh, N., Eisen, M., & Ribeiro, A. (2022) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2022.3229948 · Cited by 39
“As baselines, we compare the performance of our proposed method against three baselines: i) Full reuse (FR), where every transmitter uses Pmax; ii) Information-theoretic link scheduling (ITLinQ) [4], and iii) random-edge graph neural networks (REGNN) [10].1 We would like to point out that, while the REGNN baseline all…”
BackgroundMethodResultInfluentialSelf-citationLinkSlice: Fine-Grained Network Slice Enforcement Based on Deep Reinforcement Learning
Wang, T., Chen, S., Zhu, Y., Tang, A., & Wang, X. (2022) · IEEE Journal on Selected Areas in Communications · doi:10.1109/jsac.2022.3180776 · Cited by 37
Li, Y., Chen, Z., Wang, Y., Yang, C., Ai, B., & Wu, Y. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3218579 · Cited by 33
“Incorporating permutation equivariance into the neural network architecture can reduce the parameter space and also avoid a large number of unnecessary permuted training samples [28], [29], [31].”
BackgroundIGNNITION: Bridging the Gap between Graph Neural Networks and Networking Systems
Pujol-Perich, D., Suárez‐Varela, J., Ferriol, M., Xiao, S., Wu, B., Cabellos‐Aparicio, A., & Barlet‐Ros, P. (2021) · IEEE Network · doi:10.1109/mnet.001.2100266 · Cited by 29
“In particular, it is common to find non-standard message-passing architectures [3, 4, 6] that stem from the need to explicitly model the circular dependencies between various network elements based on the actual behavior of real network infrastructures.”
BackgroundMethodWide and Deep Graph Neural Network With Distributed Online Learning
Gao, Z., Gama, F., & Ribeiro, A. (2022) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2022.3192606 · Cited by 36
“In the context of robot communications, the GMD can be used to represent channel information [33].”
MethodSelf-citationStability of graph convolutional neural networks to stochastic perturbations
Gao, Z., Isufi, E., & Ribeiro, A. (2021) · Signal Processing · doi:10.1016/j.sigpro.2021.108216 · Cited by 28
Self-citationJoint User Scheduling and Beamforming Design for Multiuser MISO Downlink Systems
He, S., Yuan, J., An, Z., Huang, W., Huang, Y., & Zhang, Y. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3215666 · Cited by 38
“The authors of [31] proposed a random edge graph neural network (REGNN) for PA optimization on graphs formed by the interference links within WCNs.”
BackgroundGraph-Based Algorithm Unfolding for Energy-Aware Power Allocation in Wireless Networks
Li, B., Verma, G., & Segarra, S. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3204486 · Cited by 29
“cess in wireless power allocation and other communication problems [30], [31], [32], [33], [34], [35], [36], [37], [38],”
BackgroundMethodAlijoyo, F. A. (2024) · Journal of Engineering Research · doi:10.1016/j.jer.2024.01.018 · Cited by 30
Fast Power Control Adaptation via Meta-Learning for Random Edge Graph Neural Networks
Nikoloska, I., & Simeone, O. (2021) · IEEE International Workshop on Signal Processing Advances in Wireless Communications (SPAWC) · doi:10.1109/spawc51858.2021.9593131 · Cited by 28
An Unsupervised Deep Unrolling Framework for Constrained Optimization Problems in Wireless Networks
He, S., Xiong, S., An, Z., Zhang, W., Huang, Y., & Zhang, Y. (2022) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2022.3166964 · Cited by 29
“introduced random edge GNNs (REGNNs), which perform convolutions over random graphs in the wireless network and the REGNN-based resource allocation policies retain an important PE property that makes them amenable to transference to different networks [14].”
MethodZhao, B., Guo, J., & Yang, C. (2024) · IEEE Transactions on Communications · doi:10.1109/tcomm.2024.3388506 · Cited by 25
“MRT is a single-user beamforming policy and hence can be generalized to multiple users, while power allocation is easier to be generalized to users as evaluated by existing works [27], [28].”
BackgroundUnsupervised Learning for Asynchronous Resource Allocation In Ad-Hoc Wireless Networks
Wang, Z., Eisen, M., & Ribeiro, A. (2021) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp39728.2021.9414181 · Cited by 29
“Although computing gradients of (12) requires explicit knowledge of f and the distribution m(H,x), model-free learning methods such as stochastic gradient and policy gradient methods [21] are typically used to bypass this requirement by sampling the performance under different policies—see [8].”
MethodSelf-citationPower Control With QoS Guarantees: A Differentiable Projection-Based Unsupervised Learning Framework
Alizadeh, M., & Tabassum, H. (2023) · IEEE Transactions on Communications · doi:10.1109/tcomm.2023.3282220 · Cited by 29
“Based on the duality theory, [6], [7], [12] applied the Lagrangian loss function to train the DNN and parameterize both the primal and dual variables.”
MethodRuiz, L., Chamon, L. F. O., & Ribeiro, A. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9053882 · Cited by 24
“System transferability is a prevalent problem in signal processing, statistics and machine learning [1, 2].”
BackgroundSelf-citationEfficient Power Allocation Using Graph Neural Networks and Deep Algorithm Unfolding
Chowdhury, A., Verma, G., Rao, C. R., Swami, A., & Segarra, S. (2021) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp39728.2021.9415106 · Cited by 24
“On the other hand, both REGNN and IGCNet – originally designed and tested in high-noise regimes – prove to be inadequate to match the performance of WMMSE in the more challenging low-noise setting.”
BackgroundMethodInfluentialOnline Deep Neural Network for Optimization in Wireless Communications
Gao, J., Zhong, C., Li, G. Y., & Zhang, Z. (2022) · IEEE Wireless Communications Letters · doi:10.1109/lwc.2022.3149863 · Cited by 23
Graph Neural Network for Distributed Beamforming and Power Control in Massive URLLC Networks
Gu, Y., She, C., Bi, S., Quan, Z., & Vucetic, B. S. (2024) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2024.3358903 · Cited by 24
“In [22], the authors proposed a random edge graph neural network (REGNN) to solve the power allocation problem for wireless networks with multiple transceiver pairs, where the fading channels were modelled as the adjacency matrix of the graph.”
BackgroundMethodUnsupervised Resource Allocation with Graph Neural NetworksPreprint
Cranmer, M., Melchior, P., & Nord, B. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.09761 · Cited by 26
“Additional examples of GNNs used for resource allocation include [12; 9], which were applied to resources on a wireless network and focus on the edges of a GNN, whereas we use the edges merely as a way of factoring in interactions between potential subjects.”
Coordinated Beamforming for Multi-Cell ISAC Using Graph Neural Networks
Liu, X., & Fischione, C. (2025) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2025.3621437 · Cited by 24
“Assumption 1: For any set S ⊆ H of nonzero probability, there exists a nonzero probability strict subset S (cid:48) ⊂ S of lower probability, 0 E h ( I ( S (cid:48) )) E h ( I ( S )) [27].”
Privacy-Preserving Decentralized Inference With Graph Neural Networks in Wireless Networks
Lee, M., Yu, G., & Dai, H. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3279442 · Cited by 27
“…attracting increasing interest, and shown to outperform traditional model-based design paradigms in many wireless applications, such as resource allocation [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], networking [14], [15], and physical layer design [16], [17], [18], [19], [20].”
BackgroundInfluentialChien, T. V., Le, H. A., Tùng, T. H., Ngo, H. Q., & Chatzinotas, S. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2024.3423008 · Cited by 23
“Graph neural network (GNN) is a well-known approach that can explore the graph topology of radio systems to obtain a comparable performance and remarkable scalability and generalization in very large-scale systems [30], [31], [32].”
BackgroundA Study on Deep Learning for Latency Constraint Applications in Beyond 5G Wireless Systems
Sritharan, S., Weligampola, H., & Gacanin, H. (2020) · IEEE Access · doi:10.1109/access.2020.3040133 · Cited by 23
“In [26], a random edge graph neural network is proposed to parameterize the resource allocation policy which is trained using an unsupervised model-free, primal-dual learning method.”
MethodBui, V.-P., Chien, T. V., Lagunas, E., Grotz, J., Chatzinotas, S., & Ottersten, B. (2022) · IEEE Transactions on Communications · doi:10.1109/tcomm.2022.3199477 · Cited by 21
Convolutional Learning on Multigraphs
Butler, L., Parada-Mayorga, A., & Ribeiro, A. (2023) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2023.3259144 · Cited by 35
“GRAPHS have enabled the modeling of relational data in diverse scenarios [1], [2], [3], but are insufficient for representing systems with multiple edge types, such as distinct communication avenues in a social system [4] or different frequency bands in a wireless communication system [5].”
BackgroundMethodInfluentialSelf-citationA Model-Based GNN for Learning Precoding
Guo, J., & Yang, C. (2023) · IEEE Transactions on Wireless Communications · doi:10.1109/twc.2023.3336911 · Cited by 28
“All DNNs including FNNs and CNNs are possible to be generalizable to channel statistics, but only GNNs have the potential of being generalizable to problem scales owing to harnessing the PE prior [12], [13].”
BackgroundMethodMeta-Learning for Wireless Communications: A Survey and a Comparison to GNNs
Zhao, B., Wu, J., Ma, Y., & Yang, C. (2024) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2024.3380512 · Cited by 31
“Random edge GNN proposed in [24] was used to learn the policy, while each layer of the GNN was viewed as a component and can be chosen from several interchangeable modules.”
BackgroundMethodInfluentialLearning to Continuously Optimize Wireless Resource in Episodically Dynamic Environment
Sun, H., Pu, W., Zhu, M., Fu, X., Chang, T.-H., & Hong, M. (2021) · ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp39728.2021.9413503 · Cited by 21
“Different network structures, such as convolutional neural networks (CNN) [12] and graph neural networks (GNN) [14, 35], and different modeling techniques, such as reinforcement learning [36], are also studied in the literature.”
BackgroundYang, Y., Zhang, Z., Tian, Y., Jin, R., & Huang, C. (2023) · IEEE Wireless Communications · doi:10.1109/mwc.012.2200552 · Cited by 20
“Through downlink transmission, we can realize the aggregator procedure in [5, 6], or the BS-UE aggregator in [7].”
BackgroundMethodResultInfluential
Metadata
- Authors
- Eisen, M., & Ribeiro, A.
- Affiliations
- Intel (United States) (US)
- University of Pennsylvania (US)
- Venue
- IEEE Transactions on Signal Processing
- Published
- 2020
- Volume
- 68
- Pages
- 2977-2991
- ISSN
- 1053-587X (ISSN-L)
- 1941-0476
- OpenAlex
- W2971598376
- Cited by
- 327
- References
- 41
- Citations by year
- 2026: 31 · 2025: 53 · 2024: 59 · 2023: 60 · 2022: 56 · 2021: 44 · 2020: 23
- FWCI
- 15.72
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE Transactions on Signal Processing (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- ARL DCIST CRA (W911NF-17-2-0181)
- Intel Science and Technology Center for Wireless Autonomous Systems
- Topics
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Advanced Wireless Network Optimization
- Physical Sciences › Computer Science › Computer Networks and Communications › Wireless Communication Networks Research
- Physical Sciences › Computer Science › Artificial Intelligence › Advanced Graph Neural Networks
- Keywords
- wireless resource allocation, graph neural networks, permutation equivariance, model-free reinforcement learning, random graphs, convolutional neural networks
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.