BER of Massive MIMO in time-variant channels using multiplexed, superimposed and hybrid channel estimation techniques

Abstract

Massive MIMO (M-MIMO) is one of the most promising technologies towards fifth generation (5G) wireless systems. M-MIMO use a large number of antennas at the base station (BS) serving multiple user terminals (UTs). Therefore, it presents an exceptional performance in terms of the ergodic capacity and bit error rate (BER) using simpler linear detectors when time-invariant channels (TIC) and perfect channel state information (CSI) are considered. However, impairments introduced by the time-variant channel (TVC) and the imperfect channel estimation limit the M-MIMO performance. In this work, the impairments introduced by the multiplexed pilot estimation (MPE), superimposed pilot estimation (SPE), and hybrid pilot estimation (HPE), using the minimum-mean-square error (MMSE) estimator are analyzed for M-MIMO in TVCs. Furthermore, the system performance is evaluated in terms of the BER for the zero-forcing (ZF) detector. Closed-form expressions and simple lower bounds are derived. The results show that the average BER in TVCs is far from the average BER obtained for TICs. Moreover, MPE and HPE are more susceptible to the impairments of TVCs, and even increasing a considerable number of BS antennas, the BER improvement is negligible in some time-slots. Finally, the BER using SPE outperforms the other estimation techniques.