I. Introduction
It is known that the large dimensions of massive multiple-input multiple-output (MIMO) systems lead to hardening of the effective channel between the user equipment (UE) and the base station (BS), resulting in significant resilience to fading, increased spectral and energy efficiencies, as well as simpler signal processing algorithms for precoding and data detection [1]. However, these results are predicated on the availability of accurate and up to date channel state information (CSI) at the BS. The performance of massive MIMO systems is known to degrade considerably when the perfect CSI assumption is violated, due to estimation errors, pilot contamination [2], and channel aging [3]–[5]. It has recently been argued that channel aging, which manifests as a change in the propagation channel due to user mobility, limits the performance (e.g., the number of usable dimensions/users supported) of a massive MIMO system [6]. In fact, the detrimental effect of channel aging on a massive MIMO system depends heavily on the underlying communication protocol, and can only be determined via careful analysis [7].