I. Introduction
With the rapid deployment of wireless communication systems, energy efficiency (EE) becomes a key concern from the viewpoint of green communication [1], [2]. Recently, massive multiple-input–multiple-output (MIMO) systems, where a large number of antennas are deployed at the base station (BS), have attracted a great deal of research interest [3]–[12]. Massive MIMO is acknowledged as a promising technology to improve both the spectral efficiency (SE) and EE with the advantages of asymptotically negligible fast fading, noise-free channels, and arbitrarily small transmit power [13]–[15]. The major bottleneck of improving the SE in massive MIMO is the so-called pilot contamination (PC) effect, which is caused by using the nonorthogonal uplink pilot sequences at different users [3], [16]. On the other hand, a distributed antenna system (DAS), where antennas of the interested cell can either be fully distributed within the cell [17]–[19] or clustered at remote radio heads (RRHs) [20]–[23], is proven to be efficient in improving the EE and coverage by shortening the average distance between the transmitters and users, thus lowering the transmit power [24], [25]. It is expected that combining the DAS with massive MIMO by scaling up the number of antennas in the DAS, i.e., the massive DAS, can further enhance system performance [17], [20].