Noncoherent Multi-User MIMO Communications using Covariance CSIT
Christo Thomas Kurisummoottil, Wassim Tabikh, Dirk Slock, Yi, Yuan-Wu

TL;DR
This paper investigates noncoherent multi-user MIMO downlink communication using channel covariance information instead of instantaneous CSIT, proposing a pathwise covariance approach suitable for Massive MIMO and mmWave systems.
Contribution
It introduces a novel pathwise covariance CSIT framework that exploits structured low-rank covariances, enabling distributed beamforming with limited CSIT and analyzing spectral efficiency trade-offs.
Findings
pwCSIT-based designs achieve near iCSIT spectral efficiency with less CSIT overhead
The pathwise approach enables distributed beamforming with local covariance information
Spectral efficiency analysis in low and high SNR regimes shows comparable performance
Abstract
The Multi-User downlink, particularly in a Multi-Cell Massive MIMO setting, requires enormous amounts of instantaneous CSIT (Channel State Information at the Transmitter(s)), iCSIT. Here we focus on exploiting channel covariance CSIT (coCSIT) only. In particular multipath induced structured low rank covariances are considered that arise in Massive MIMO and mmWave settings, which we call pathwise CSIT (pwCSIT). The resulting non-Kronecker MIMO channel covariance structures lead to a split between the roles of transmitters and receivers in MIMO systems. For the beamforming optimization, we consider a minorization approach applied to the Massive MIMO limit of the Expected Weighted Sum Rate. Simulations indicate that the pwCSIT based designs may lead to limited spectral efficiency loss compared to iCSIT based designs, while trading fast fading CSIT for slow fading CSIT. We also point out…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
