Abstract:Deep learning has recently emerged as a disruptive technology to solve challenging radio resource management problems in wireless networks. However, the neural network architectures adopted by existing works suffer from poor scalability, generalization, and lack of interpretability. A long-standing approach to improve scalability and generalization is to incorporate the structures of the target task into the neural network architecture. In this paper, we propose to apply graph neural networks (GNNs) to solve large-scale radio resource management problems, supported by effective neural network architecture design and theoretical analysis. Specifically, we first demonstrate that radio resource management problems can be formulated as graph optimization problems that enjoy a universal permutation equivariance property. We then identify a class of neural networks, named \emph{message passing graph neural networks} (MPGNNs). It is demonstrated that they not only satisfy the permutation equivariance property, but also can generalize to large-scale problems while enjoying a high computational efficiency. For interpretablity and theoretical guarantees, we prove the equivalence between MPGNNs and a class of distributed optimization algorithms, which is then used to analyze the performance and generalization of MPGNN-based methods. Extensive simulations, with power control and beamforming as two examples, will demonstrate that the proposed method, trained in an unsupervised manner with unlabeled samples, matches or even outperforms classic optimization-based algorithms without domain-specific knowledge. Remarkably, the proposed method is highly scalable and can solve the beamforming problem in an interference channel with $1000$ transceiver pairs within $6$ milliseconds on a single GPU.

Graph Neural Network Based Resource Allocation for Full-duplex Device-to-Device Communication

Deep Learning Based Resource Allocation for Full-duplex Device-to-Device Communication

Deep Learning Based Resource Allocation for Full-Duplex-Enabled Two-Way Device-To-Device Communications

Dynamic Resource Allocation for Device‐to‐Device Communication Underlaying Cellular Networks

Knowledge-Driven Resource Allocation for D2D Networks: A WMMSE Unrolled Graph Neural Network Approach

Graph-based Resource Allocation for Device-to-device Communications Aided Cellular Network

Graph Neural Networks and Deep Reinforcement Learning Based Resource Allocation for V2X Communications

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

Machine Learning-Based Resource Optimization for D2D Communication Underlaying Networks

Hypergraph Based Resource Allocation for Cross-Cell Device-to-device Communications

Graph neural network-based scheduling for multi-UAV-enabled communications in D2D networks

Hybrid Frequency Band Communication Scheme in Device to Device Based on Convolutional Neural Network

Graph Coloring Based Resource Sharing (GCRS) Scheme for D2D Communications Underlaying Full-Duplex Cellular Networks

Joint Mode Selection and Resource Allocation for Device-to-Device Communications

Interference-aware Graph Based Resource Sharing for Device-to-device Communications Underlaying Cellular Networks

Graph Neural Networks for Scalable Radio Resource Management: Architecture Design and Theoretical Analysis

Joint Admission and Power Control for Massive Connections via Graph Neural Network

Interference Graph-Based Resource Allocation (ingra) for D2D Communications Underlaying Cellular Networks

Deep-Learning-Based Resource Allocation for Time-Sensitive Device-to-Device Networks

Graph-based resource allocation for D2D communications underlaying cellular networks

A Two-Step Resource Allocation Algorithm for D2D Communication in Full Duplex Cellular Network.