RISnet: A Domain-Knowledge Driven Neural Network Architecture for RIS Optimization with Mutual Coupling and Partial CSI

2024-03-06

IEEE Transactions on Wireless Communications

AI Summary

This work addresses three key challenges in jointly optimizing base station precoding and reconfigurable intelligent surface (RIS) configuration for RIS-aided SDMA systems: (i) modeling electromagnetic mutual coupling among RIS elements, (ii) scalability limitations with thousand-element RISs, and (iii) heavy reliance on perfect channel state information (CSI). To this end, we propose a novel machine learning paradigm infused with electromagnetic prior knowledge. Our method introduces RISnet—a permutation-invariant neural network embedding a mutual-coupling-aware channel model and an unsupervised training scheme—and establishes a hybrid optimization framework that synergizes ML-driven RIS configuration with closed-form precoding. Experiments demonstrate robust performance under partial CSI, significantly reduced computational complexity versus conventional approaches, real-time configurability for RISs exceeding 1,000 elements, and superior generalization capability and energy efficiency compared to generic ML baselines.

Technology Category

• Machine Learning: Neuro-Symbolic Learning
• Intelligent Robots: Learning & Optimization for ROB
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Security and Privacy: Security and privacy of machine learning and AI applications
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Space-division multiple access (SDMA) plays an important role in modern wireless communications. Its performance depends on the channel properties, which can be improved by reconfigurable intelligent surfaces (RISs). In this work, we jointly optimize SDMA precoding at the base station (BS) and RIS configuration. We tackle difficulties of mutual coupling between RIS elements, scalability to more than 1000 RIS elements, and high requirement for channel estimation. We first derive an RIS-assisted channel model considering mutual coupling, then propose an unsupervised machine learning (ML) approach to optimize the RIS with a dedicated neural network (NN) architecture RISnet, which has good scalability, desired permutation-invariance, and a low requirement for channel estimation. Moreover, we leverage existing high-performance analytical precoding scheme to propose a hybrid solution of ML-enabled RIS configuration and analytical precoding at BS. More generally, this work is an early contribution to combine ML technique and domain knowledge in communication for NN architecture design. Compared to generic ML, the problem-specific ML can achieve higher performance, lower complexity and permutation-invariance.

Problem

• Optimize SDMA precoding and RIS configuration jointly.
• Address mutual coupling and scalability in RIS elements.
• Reduce channel estimation requirements using RISnet architecture.

Innovation

• Unsupervised ML optimizes RIS with RISnet
• Hybrid ML-analytical precoding enhances BS performance
• Domain-knowledge driven NN ensures scalability, low complexity