Table 7 Channel hardening in cell-free massive MIMO

[62]

This survey dwells on incorporating channel hardening in CF-MIMO using stochastic geometry while evaluating potential constraints to its practical implementation. A homogenous PPP is explored to model the AP distribution. In addition, the effect of channel hardening on AP density is evaluated.

▪ The study reveals that channel hardening in CF-mMIMO mainly depends on the number of antennas per AP and the pathloss exponent of the propagation environment.

▪ A synergy between smaller pathloss exponent and multiple antennas per AP provides a modest level of channel hardening.

▪ The capacity bounds obtained from cellular massive MIMO are not so tight in CF-mMIMO.

▪ Channel hardening comes at the cost of reduced macro-diversity, which is undesirable in practice.

2017

[64]

The authors considered the channel hardening effect extensively in CF massive MIMO under practical operational conditions. Closed-form expressions are presented for the hardening coefficients, taking cognizance of pilot contamination, power loading, and MMSE channel estimation.

▪ Channel hardening is substantially improved with the NCB precoder compared to the CB scheme.

▪ The effects of spatial correlation at the APs are not considered.

▪ The system is largely affected by pilot contamination, given a small coverage area.

2020

[133]

An enhanced normalized CB (ECB) precoding is proposed to boost the channel hardening effect in CF massive MIMO. An exact closed-form expression for the achievable DL SE concerning the pilot contamination, channel estimation errors, lack of CSI at the user side, along with an optimal max-min fairness power allocation scheme, is presented.

▪ The effective channel is nearly deterministic, thus, validating the effectiveness of the proposed ECB.

▪ The DL SE is significantly improved compared to the case with CB.

▪ The proposed scheme does not support interference suppression.

▪ Acquiring CSI for the users is quite challenging.

2021