Abstract: Beamforming design for intelligent reflecting surface (IRS)-assisted multi-user communication (IRS-MUC) systems critically depends on the acquisition of accurate channel state information (CSI). However, channel estimation (CE) in IRS-MUC systems causes a large signaling overhead for training due to the large number of IRS elements. In this paper, taking into account user mobility, we adopt a deep learning (DL) approach to implicitly learn the historical line-of-sight (LoS) channel features and predict the IRS phase shifts to be adopted for the next time slot for maximization of the weighted sum-rate (WSR) of the IRS-MUC system. With the proposed predictive approach, we can avoid full-scale CSI estimation and facilitate low-dimensional CE for transmit beamforming design such that the signaling overhead is reduced by a scale of $\frac{1}{N}$, where $N$ is the number of IRS elements. To this end, we first develop a universal DL-based predictive beamforming (DLPB) framework featuring a two-stage predictive-instantaneous beamforming mechanism. As a realization of the developed framework, a location-aware convolutional long short-term memory (CLSTM) graph neural network (GNN) is developed to facilitate effective predictive beamforming at the IRS, where a CLSTM module is first adopted to exploit the spatial and temporal features of the considered channels and a GNN is then applied to empower the designed neural network with high scalability and generalizability. Furthermore, in the second stage, based on the predicted IRS phase shifts, an instantaneous CSI-aware fully-connected neural network is designed to optimize the transmit beamforming at the access point. Simulation results demonstrate that the proposed framework not only achieves a better WSR performance and requires a lower CE overhead compared with state-of-the-art benchmarks, but also is highly scalable in the numbers of users.

Stratified Federated Learning-Based Beamforming Design for Heterogeneous IRSs-Aided Networks.

A Distributed Machine Learning-Based Approach for IRS-Enhanced Cell-Free MIMO Networks

Design of Beamforming Algorithm for Ultra-reliable and Low-latency Communication in Heterogeneous Networks Based on IRS Assistance

Distributed Rate Optimization for Intelligent Reflecting Surface with Federated Learning.

Heterogeneous Multi-Agent Reinforcement Learning for Joint Active and Passive Beamforming in IRS Assisted Communications

A Robust Deep Learning-Based Beamforming Design for RIS-Assisted Multiuser MISO Communications with Practical Constraints

Hybrid Beamforming for Distributed Intelligent Reflecting Surfaces-Aided Systems

Enabling Efficient Scheduling Policy in Intelligent Reflecting Surface Aided Federated Learning

Two Power Allocation and Beamforming Strategies for Active IRS-aided Wireless Network via Machine Learning

Amplitude-Dependent Phase-Gradient Directional Beamforming for IRS: A Scalable Optimization Framework

Scalable Predictive Beamforming for IRS-Assisted Multi-User Communications: A Deep Learning Approach

Deep-Unfolding Beamforming for Intelligent Reflecting Surface assisted Full-Duplex Systems

Distributed Beam Training for Intelligent Reflecting Surface Enabled Multi-Hop Routing

Scalable Predictive Beamforming for IRS-Assisted Multi-User Communications: A Deep Learning Approach

Deep Learning-Empowered Predictive Beamforming for IRS-Assisted Multi-User Communications.

Label-free Optimization for Passive Beamforming in IRS-assisted MISO Systems

QoS-Constrained Federated Learning Empowered by Intelligent Reflecting Surface

Base Station Beamforming Design in Near-field XL-IRS Beam Training

Beamforming design via machine learning in intelligent reflecting surface-aided wireless communication

Passive Beamforming for IRS Aided Wireless Networks

Adaptive Beamforming for Non-Line-of-Sight IRS-Assisted Communications Without CSI