Downlink Performance of CF Massive MIMO Under Wireless-Based Fronthaul Network

Abstract

This paper investigates the downlink (DL) performance of cell-free (CF) massive multiple-input multiple-output (mMIMO) systems under a wireless mMIMO-based fronthaul network operation. Particularly, we consider multiple edge-cloud processors (ECP)s serving access points (AP)s using one of three possible fronthaul network operations, namely, microwave, millimeter wave (mmWave), or hybrid microwave/mmWave. Under each fronthaul network operation, we analyze the achievable DL data rates for two different microwave-based operations of the access link (APs-users), namely, distributed and centralized operations, while assuming APs with/without decoding capabilities. In the distributed operation, APs are responsible for performing both channel estimation and DL data precoding tasks, whereas ECPs are the responsible entities for carrying out such tasks in the centralized counterpart. Our results show that the integration between the centralized access link operation and the hybrid-based fronthaul network provides the highest DL data rates when APs are empowered with decoding capabilities. However, integrating the distributed access link operation with the microwave-based fronthaul network achieves ultimate performance when APs are not supported with decoding capabilities. Interestingly, we reveal that APs with low fronthaul capacities dominantly control the preferred network configuration in-terms of the densities and the number of deployed antennas for both APs and ECPs.