Heterogeneous graph neural network for power allocation in multicarrier-division duplex cell-free massive MIMO systems

Li, Bohan; Yang, Lie‐Liang; Maunder, Robert G.; Sun, Songlin; Xiao, Ping

doi:10.48550/arxiv.2205.00546

Cited by 2 publications

(6 citation statements)

References 38 publications

(59 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The feasible region of optimization variables can be highly stringent to satisfy practical network conditions under resource constraints and provide robust and reliable services for ubiquitous networking. For example, the spectral-efficiency maximization problem in heterogeneous networks involves lots of non-convex constraints to support different transmission types (i.e., uplink and downlink transmission constraints) [10]. (3) The optimization problems in 6G wireless networks usually involve real-time network-dependent parameters, such as the network structure, channel state information (CSI), traffic condition, etc.…”

Section: A Large-scale Optimization For 6gmentioning

confidence: 99%

“…) , (10) where PL j∈N (k) (•) denotes the pooling function used for aggregating the outputs of the nodes in N (k), N (k) denotes the set of neighboring nodes of node k, f…”

Section: A Principles Of Graph Neural Networkmentioning

confidence: 99%

“…can be modeled as different types of nodes connected by different types of edges (e.g., transmission edges and inference edges) to indicate a more complex wireless communication network. A heterogeneous graph neural network (HGNN) was firstly proposed in [10] to characterize two different types of nodes (access points (APs) and mobile stations (MSs)) and various types of edges between nodes (e.g., uplink/downlink transmission path and inter-AP/inter-MS interference path) in distributed cell-free massive systems. To address the impact of heterogeneous nodes, an adaptive node embedding layer was proposed, where the node features of AP and MS are transformed by two embedding matrices to handle the varying input feature dimensions before the process of GNN layers.…”

Section: E Application 3: Graph Neural Network In Distributed Systemsmentioning

confidence: 99%

See 2 more Smart Citations

Machine Learning for Large-Scale Optimization in 6G Wireless Networks

Shi¹,

Lian²,

Shi³

et al. 2023

Preprint

View full text Add to dashboard Cite

The sixth generation (6G) wireless systems are envisioned to enable the paradigm shift from "connected things" to "connected intelligence", featured by ultra high density, largescale, dynamic heterogeneity, diversified functional requirements and machine learning capabilities, which leads to a growing need for highly efficient intelligent algorithms. The classic optimization-based algorithms usually require highly precise mathematical model of data links and suffer from poor performance with high computational cost in realistic 6G applications. Based on domain knowledge (e.g., optimization models and theoretical tools), machine learning (ML) stands out as a promising and viable methodology for many complex large-scale optimization problems in 6G, due to its superior performance, generalizability, computational efficiency and robustness. In this paper, we systematically review the most representative "learning to optimize" techniques in diverse domains of 6G wireless networks by identifying the inherent feature of the underlying optimization problem and investigating the specifically designed ML frameworks from the perspective of optimization. In particular, we will cover algorithm unrolling, learning to branch-andbound, graph neural network for structured optimization, deep reinforcement learning for stochastic optimization, as well as endto-end learning for semantic optimization, for solving challenging large-scale optimization problems arising from various important wireless applications. To enable ML implementation in distributed wireless networks across massive number of end devices, federated learning for distributed optimization will further be presented. Through the in-depth discussion, we shed light on the excellent performance of ML-based optimization algorithms with respect to the classical methods, and provide insightful guidance to develop advanced ML techniques in 6G networks. Neural network design, theoretical tools of different ML methods,

show abstract

Section: A Large-scale Optimization For 6gmentioning

confidence: 99%

“…) , (10) where PL j∈N (k) (•) denotes the pooling function used for aggregating the outputs of the nodes in N (k), N (k) denotes the set of neighboring nodes of node k, f…”

Section: A Principles Of Graph Neural Networkmentioning

confidence: 99%

Section: E Application 3: Graph Neural Network In Distributed Systemsmentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning for Large-Scale Optimization in 6G Wireless Networks

Shi¹,

Lian²,

Shi³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Comprehensive simulations demonstrate that the MPGNNs have a similar performance to the weighted sum MSE minimization (WMMSE) algorithm [2] with less computational complexity and the model has good generalisation capacities. Heterogeneous GNNs with a novel graph representation are proposed in [6] and [7] to allocate power in device-to-device (D2D) and cell-free massive Multiple-Input-Multiple-Output (MIMO) networks, respectively.…”

mentioning

confidence: 99%

“…However, as a promising technique to provide the potential of doubling the capacity compared to conventional HD transmission [8], full-duplex (FD) transmission has drawn much attention recently. To the best of our knowledge, current GNN-based methods in power allocation only focus on HD transmission, and only a few papers focus on FD transmission such as [7].…”

mentioning

confidence: 99%

Graph Neural Networks for Power Allocation in Wireless Networks with Full Duplex Nodes

Chen¹,

Zhu²,

Evans³

2023

Preprint

View full text Add to dashboard Cite

Due to mutual interference between users, power allocation problems in wireless networks are often non-convex and computationally challenging. Graph neural networks (GNNs) have recently emerged as a promising approach to tackling these problems and an approach that exploits the underlying topology of wireless networks. In this paper, we propose a novel graph representation method for wireless networks that include full-duplex (FD) nodes. We then design a corresponding FD Graph Neural Network (F-GNN) with the aim of allocating transmit powers to maximise the network throughput. Our results show that our F-GNN achieves state-of-art performance with significantly less computation time. Besides, F-GNN offers an excellent trade-off between performance and complexity compared to classical approaches.We further refine this trade-off by introducing a distance-based threshold for inclusion or exclusion of edges in the network. We show that an appropriately chosen threshold reduces required training time by roughly 20% with a relatively minor loss in performance.

show abstract

Heterogeneous graph neural network for power allocation in multicarrier-division duplex cell-free massive MIMO systems

Cited by 2 publications

References 38 publications

Machine Learning for Large-Scale Optimization in 6G Wireless Networks

Machine Learning for Large-Scale Optimization in 6G Wireless Networks

Graph Neural Networks for Power Allocation in Wireless Networks with Full Duplex Nodes

Contact Info

Product

Resources

About