User-Centric Massive MIMO Systems with Hardening-based Clusterization

Abstract:
When speaking of interference management and avoidance for future wireless communications, massive multipleinput multiple-output (MIMO) cell-free (CF) systems are becoming a key technology to be considered in upcoming standards. Some different approaches based on the original proposal have arisen in the past few years for either dealing with the limited fronthaul links capacity or making the whole system scalable for real deployments. The user-centric (UC) massive MIMO approach limits the number of access points (AP) serving a particular user. Some performance is lost this way, however, the fronthaul requirements seem much more feasible in terms of capacity demands. In this paper, we focus on a similar approach giving a novel criteria for the AP-selection based on a channel hardening metric. We ensure that the number of APs serving a user matches a hardening target threshold, making the achievable lower bound rates much closer to the upper bound while reducing fronthaul capacity demands. We show that our proposed non-arbitrary AP-selection strategy performs fairly good when setting an appropriate hardening target if compared with the pure CF approach and outperforms the arbitrary UC approach.