Abstract:Future wireless systems are trending towards higher carrier frequencies that offer larger communication bandwidth but necessitate the use of large antenna arrays. Existing signal processing techniques for channel estimation do not scale well to this "high-dimensional" regime in terms of performance and pilot overhead. Meanwhile, training deep learning based approaches for channel estimation requires large labeled datasets mapping pilot measurements to clean channel realizations, which can only be generated offline using simulated channels. In this paper, we develop a novel unsupervised over-the-air (OTA) algorithm that utilizes noisy received pilot measurements to train a deep generative model to output beamspace MIMO channel realizations. Our approach leverages Generative Adversarial Networks (GAN), while using a conditional input to distinguish between Line-of-Sight (LOS) and Non-Line-of-Sight (NLOS) channel realizations. We also present a federated implementation of the OTA algorithm that distributes the GAN training over multiple users and greatly reduces the user side computation. We then formulate channel estimation from a limited number of pilot measurements as an inverse problem and reconstruct the channel by optimizing the input vector of the trained generative model. Our proposed approach significantly outperforms Orthogonal Matching Pursuit on both LOS and NLOS channel models, and EM-GM-AMP -- an Approximate Message Passing algorithm -- on LOS channel models, while achieving comparable performance on NLOS channel models in terms of the normalized channel reconstruction error. More importantly, our proposed framework has the potential to be trained online using real noisy pilot measurements, is not restricted to a specific channel model and can even be utilized for a federated OTA design of a dataset generator from noisy data.

Channel Estimation for Cell-Free Massive MIMO Using Conditional GAN.

GAN-based Channel Estimation for Massive MIMO System

Block Expectation Propagation for Downlink Channel Estimation in Massive MIMO Systems.

ChannelGAN: Deep Learning-Based Channel Modeling and Generating

GAN Based Near-Field Channel Estimation for Extremely Large-Scale MIMO Systems

Channel Estimation for Wireless Cellular Systems with Massive Linear Receiving Antennas

Over-the-Air Design of GAN Training for mmWave MIMO Channel Estimation

Improved GAN-Based Channel Estimation Methods for MIMO OFDM Systems

RIDNet Assisted cGAN Based Channel Estimation for One-Bit ADC mmWave MIMO Systems

A GAN-Based Channel Estimation Method for MIMO OFDM Systems

An Effective Generative Model Based Channel Estimation Method With Reduced Overhead

Federated Generative Adversarial Networks Based Channel Estimation

Channel Estimation for mmWave Massive MIMO Systems Using Graph Neural Networks

Tcgtr-Based Channel Estimation For Cooperative Multi-Cell Large Scale Antenna Systems

Semi-blind Channel Estimation Leveraging Frequency Correlation

Channel Estimation for Intelligent Reflecting Surface Aided Wireless Communications Using Conditional GAN

Deep Neural Network Based Channel Estimation for Massive MIMO-OFDM Systems With Imperfect Channel State Information

Pay Less but Get More: A Dual-Attention-based Channel Estimation Network for Massive MIMO Systems with Low-Density Pilots.

Distributed Conditional Generative Adversarial Networks (GANs) for Data-Driven Millimeter Wave Communications in UAV Networks

Federated Generative-Adversarial-Network-Enabled Channel Estimation

Channel Estimation for Fiber-Terahertz Communication Based on Conditional Generative Adversarial Networks