Contents I. Introduction
Massive multiple-input multiple-output (MIMO) is one of the key technology of fifth generation (5G) to solve the 10,000- fold capacity increase compared to long term evolution (LTE) in use cases such as enhanced Mobile-BroadBand (eMBB). As an intermediate step LTE Release 13 introduced a study item on full-dimension MIMO (FD-MIMO) [1] assuming 16 transmit antenna ports, while Release 14 already covers 32 antenna ports [2]. In the first 5G version, Release 15 also called new radio (NR), 3rd generation partnership project (3GPP) considers 64 antenna ports at the base station (BS) side [3]. Massive MIMO capacity gains are mainly achieved by spatial multiplexing or short multi-user MIMO. Massive MIMO literature often considers uncorrelated or orthogonal channels among multiple users. Consequently, the channel rank and subsequent capacity is limited by number of users and transmit power. However, it is known that the assumption of orthogonal channels often does not hold, especially for centralized arrays and antenna numbers as supported in 5G. Therefore, urban macro channel measurements have been performed in Berlin city covering the frequency range from 3.55 GHz to 3.8 GHz that is used for 5G deployment all over the word, e.g. in Germany [4] or the United States [5]. In the evaluation of these measurement we focus on the spatial correlation of multi-user channels and its impact on throughput.
