Conjugate Beamforming with Fractional-Exponent Normalization and Scalable Power Control in Cell-Free Massive MIMO

TL;DR

This paper introduces a novel conjugate beamforming technique with fractional-exponent normalization and a scalable power control strategy for cell-free massive MIMO, improving efficiency and reducing power consumption.

Contribution

It derives a generalized spectral efficiency expression for CF-mMIMO with fractional-exponent normalization and proposes a low-complexity power control method based on long-term fading.

Findings

Performance close to max-min power control with lower complexity

Reduced power consumption compared to traditional methods

Effective spectral efficiency with fractional-exponent normalization

Abstract

This paper considers a cell-free massive MIMO (CF-mMIMO) system using conjugate beamforming (CB) with fractional-exponent normalization. Assuming independent Rayleigh fading channels, a generalized closed-form expression for the achievable downlink spectral efficiency is derived, which subsumes, as special cases, the spectral efficiency expressions previously reported for plain CB and its variants, i.e. normalized CB and enhanced CB. Downlink power control is also tackled, and a reduced-complexity power allocation strategy is proposed, wherein only one coefficient for access point (AP) is optimized based on the long-term fading realizations. Numerical results unveil the performance of CF-mMIMO with CB and fractional-exponent normalization, and show that the proposed power optimization rule incurs a moderate performance loss with respect to the traditional max-min power control rule, but with lower complexity and much smaller overall power consumption.