An Unsupervised Learning-Based Approach for Symbol-Level-Precoding

TL;DR

This paper introduces an unsupervised deep learning framework for symbol-level precoding that minimizes transmit power while satisfying SINR constraints, outperforming traditional methods in speed and near-optimality.

Contribution

It presents a novel unsupervised neural network architecture for symbol-level precoding based on unfolding an interior point method, which is a new approach in this domain.

Findings

Outperforms conventional block-level precoding in power efficiency.

Achieves near-optimal performance faster than traditional optimization.

Effectively exploits known interference for power minimization.

Abstract

This paper proposes an unsupervised learning-based precoding framework that trains deep neural networks (DNNs) with no target labels by unfolding an interior point method (IPM) proximal `log' barrier function. The proximal `log' barrier function is derived from the strict power minimization formulation subject to signal-to-interference-plus-noise ratio (SINR) constraint. The proposed scheme exploits the known interference via symbol-level precoding (SLP) to minimize the transmit power and is named strict Symbol-Level-Precoding deep network (SLP-SDNet). The results show that SLP-SDNet outperforms the conventional block-level-precoding (Conventional BLP) scheme while achieving near-optimal performance faster than the SLP optimization-based approach.