Uplink Massive MIMO for Channels with Spatial Correlation

TL;DR

This paper analyzes uplink massive MIMO systems with spatially correlated channels, extending LSFP techniques to improve spectral efficiency and reduce pilot contamination, resulting in significant data rate enhancements.

Contribution

It extends Large Scale Fading Precoding to correlated channels and derives SINR expressions that enable effective power optimization in finite antenna regimes.

Findings

SINR expressions depend only on slow fading components.

Power optimization algorithms significantly increase data rates.

Spatial correlation can be exploited to mitigate pilot contamination.

Abstract

A massive MIMO system entails a large number of base station antennas M serving a much smaller number of users. This leads to large gains in spectral and energy efficiency compared with other technologies. As the number of antennas M grows, the performance of such systems gets limited by pilot contamination interference. Large Scale Fading Precoding/Postcoding (LSFP) was proposed in literature for mitigation of pilot contamination. It was shown recently that in channels without spatial correlation (uncorrelated base station antennas) LSFP leads to large spectral-efficiency gains. Also, recently, it was observed that if a channel has spatial correlation, then one can use this correlation to drastically reduce the pilot contamination interference in the asymptotic regime as M tends to infinity. In this work, we analyze the performance of Uplink (UL) transmission of massive MIMO systems with finitely many antennas M for channels with spatial correlation. We extend the idea of LSFP to correlated channel models and derive SINR expressions that depend only on slow fading channel components for such systems with and without LSFP. These simple expressions lead us to simple algorithms for transmit power optimization. As a result, we obtain a multi-fold increase in data transmission rates.