A CNN-based End-to-End Learning for RIS-assisted Communication System

2025-03-18

AI Summary

Conventional RIS-aided communication systems suffer from performance bottlenecks due to the decoupled, stepwise optimization of transmitter, receiver, and RIS phase configurations. Method: This paper proposes an end-to-end trainable CNN autoencoder architecture—the first to jointly learn channel coding/decoding, modulation/demodulation, channel estimation, and RIS phase control. A differentiable RIS channel model is introduced to tightly integrate physical-layer signal processing with intelligent reflecting surface control, thereby departing from traditional layered design paradigms. Contribution/Results: Experimental results demonstrate that, at identical SNR levels, the proposed method achieves a lower bit error rate (BER) than the theoretically optimal sequential RIS design. This validates both the feasibility and superiority of end-to-end learning for 6G RIS-assisted wireless communications.

Technology Category

• Search and Optimization: Learning to Search
• Intelligent Robots: Behavior Learning & Control
• Computer Vision: Learning & Optimization for CV

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Reconfigurable intelligent surface (RIS) is an emerging technology that is used to improve the system performance in beyond 5G systems. In this letter, we propose a novel convolutional neural network (CNN)-based autoencoder to jointly optimize the transmitter, the receiver, and the RIS of a RIS-assisted communication system. The proposed system jointly optimizes the sub-tasks of the transmitter, the receiver, and the RIS such as encoding/decoding, channel estimation, phase optimization, and modulation/demodulation. Numerically we have shown that the bit error rate (BER) performance of the CNN-based autoencoder system is better than the theoretical BER performance of the RIS-assisted communication systems.

Problem

• Optimize RIS-assisted communication system performance
• Jointly optimize transmitter, receiver, and RIS tasks
• Improve bit error rate using CNN-based autoencoder

Innovation

• CNN-based autoencoder for RIS-assisted communication
• Joint optimization of transmitter, receiver, RIS
• Improved BER performance over theoretical systems