Model-Driven Deep Learning for Massive Multiuser MIMO Constant Envelope Precoding

TL;DR

This paper introduces a model-driven deep learning approach for constant envelope precoding in massive MIMO systems, reducing computational complexity while maintaining interference suppression.

Contribution

It proposes a novel deep learning network that unfolds and parameterizes the conjugate gradient algorithm for efficient CE precoding.

Findings

Outperforms traditional algorithms in interference suppression

Reduces computational overhead significantly

Learns optimal search parameters through unsupervised training

Abstract

Constant envelope (CE) precoding design is of great interest for massive multiuser multi-input multi-output systems because it can significantly reduce hardware cost and power consumption. However, existing CE precoding algorithms are hindered by excessive computational overhead. In this letter, a novel model-driven deep learning (DL)-based network that combines DL with conjugate gradient algorithm is proposed for CE precoding. Specifically, the original iterative algorithm is unfolded and parameterized by trainable variables. With the proposed architecture, the variables can be learned efficiently from training data through unsupervised learning approach. Thus, the proposed network learns to obtain the search step size and adjust the search direction. Simulation results demonstrate the superiority of the proposed network in terms of multiuser interference suppression capability and computational overhead.