Transformer-Driven Neural Beamforming with Imperfect CSI in Urban Macro Wireless Channels

2025-04-15

AI Summary

This work addresses the beamforming problem for macro-cell multi-user MIMO systems in dense urban environments under imperfect channel state information (CSI). We propose an unsupervised, end-to-end neural network-based beamforming (NNBF) framework. Our method innovatively integrates depthwise separable convolutions with Transformer-based self-attention to enable lightweight, channel-aware beamweight generation—without requiring ideal CSI labels or any supervised training signals. Instead, it is trained solely via self-supervision using sum-rate and block error rate (BLER) as optimization objectives. Experimental results under realistic urban channel models demonstrate that the proposed NNBF achieves a 23.6% gain in sum-rate, a 41.2% reduction in BLER, and a 19.8% improvement in spectral efficiency compared to conventional zero-forcing (ZF) and minimum mean square error (MMSE) beamformers—effectively balancing throughput enhancement and communication reliability.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Search and Optimization: Non-convex Optimization
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

The literature is abundant with methodologies focusing on using transformer architectures due to their prominence in wireless signal processing and their capability to capture long-range dependencies via attention mechanisms. In particular, depthwise separable convolutions enhance parameter efficiency for the process of high-dimensional data characteristics of MIMO systems. In this work, we introduce a novel unsupervised deep learning framework that integrates depthwise separable convolutions and transformers to generate beamforming weights under imperfect channel state information (CSI) for a multi-user single-input multiple-output (MU-SIMO) system in dense urban environments. The primary goal is to enhance throughput by maximizing sum-rate while ensuring reliable communication. Spectral efficiency and block error rate (BLER) are considered as performance metrics. Experiments are carried out under various conditions to compare the performance of the proposed NNBF framework against baseline methods zero-forcing beamforming (ZFBF) and minimum mean square error (MMSE) beamforming. Experimental results demonstrate the superiority of the proposed framework over the baseline techniques.

Problem

• Develop transformer-based beamforming for imperfect CSI in urban macro channels
• Enhance throughput and sum-rate in MU-SIMO systems using deep learning
• Improve spectral efficiency and BLER compared to ZFBF and MMSE methods

Innovation

• Transformer and depthwise separable convolutions integration
• Unsupervised deep learning for imperfect CSI
• Enhances throughput in urban MU-SIMO systems