Turbo-AI: Iterative Machine Learning Based Channel Estimation for 2D Massive Arrays

TL;DR

This paper introduces Turbo-AI, an iterative machine learning approach for efficient 2D massive array channel estimation, leveraging subspace training, noise variance reduction, and universal training to improve accuracy and reduce complexity.

Contribution

The paper proposes Turbo-AI, a novel iterative ML-based channel estimation method with subspace training and universal training for 2D massive arrays, enhancing efficiency and accuracy.

Findings

Turbo-AI approaches the genie-aided bound at low SNR.

Subspace training reduces computational complexity significantly.

Universal training enables a single NN to operate across various SNRs and angles.

Abstract

Recently, Machine Learning (ML) is recognized as an effective tool for wireless communications and plays an evolutionary role to enhance Physical Layer (PHY) of 5th Generation (5G) and Beyond 5G (B5G) systems. In this paper, we focus on the ML-based channel estimation for 2- Dimensional (2D) massive antenna arrays. Due to the extremely high computational requirement for 2D arrays with Ordinary Training, we exploit 2D Kronecker covariance model to perform Subspace Training for vertical and horizontal spatial domain independently, which achieves a complexity cost saving factor $O(M^4N^4)/O(MN^4 + NM^4)$ for ML with an $M \times N$ 2D array. Furthermore, we propose an iterative training approach, referred to as Turbo-AI. Along with Subspace Training, the new approach can monotonically reduce the effective variance of additive noise of the observation, and update the Neural Network (NN) models by re-training. Furthermore, we propose a concept, named Universal Training. It allows to use one NN for a wide range of Signal-to-Noise-Ratio (SNR) operation points and spatial angles, which can greatly simplify Turbo-AI usage. Numerical results exhibit that Turbo-AI can tightly approach the genie-aided channel estimation bound, especially at low SNR.