Compressed CSI Feedback With Learned Measurement Matrix for mmWave Massive MIMO

Abstract

A major challenge to implement the compressed sensing method for channel state information (CSI) acquisition lies in the design of a well-performed measurement matrix to reduce the dimension of sparse channel vectors. The widely adopted randomized measurement matrices drawn from Gaussian or Bernoulli distribution are not optimal. To tackle this problem, we propose a fully data-driven approach to optimize the measurement matrix for beamspace channel compression, and this method trains a mathematically interpretable autoencoder constructed according to the iterative solution of sparse recovery. The obtained measurement matrix can achieve near perfect CSI recovery with fewer measurements, thus the feedback overhead can be substantially reduced.

arXiv comment4 pages, 3 figures

References in corpus (3)

• An Overview of Signal Processing Techniques for Millimeter Wave MIMO Systems
• Compressed Sensing Based Multi-User Millimeter Wave Systems: How Many Measurements Are Needed?
• Learning a Compressed Sensing Measurement Matrix via Gradient Unrolling

Cited by in corpus (3)

• Convolutional Neural Network based Multiple-Rate Compressive Sensing for Massive MIMO CSI Feedback: Design, Simulation, and Analysis
• A Survey on Deep-Learning based Techniques for Modeling and Estimation of MassiveMIMO Channels
• Deep Learning for 1-Bit Compressed Sensing-based Superimposed CSI Feedback