Search

This paper investigates the downlink (DL) rate of multicell massive multiuser multiple-input and multiple-output systems over Ricean fading channels that takes into account channel estimation errors. To acquire channel state information at all users, beamforming training (BT) is examined. Considering both maximum-ratio transmission (MRT) and zero forcing, this paper derives closed-form expressions on the lower bound of the achievable rates for two cases, with or without BT. With the obtained expressions, Bernoulli's inequality is invoked to find the ranges for the length of DL pilots such that the sum spectral efficiency of the scheme with BT is superior to that of the scheme without BT, and vice versa. Various power scaling laws concerning DL data and pilot transmit powers and uplink pilot transmit power are analyzed. Numerical results corroborate the accuracy of the closed-form expressions. In particular, the results show that employing BT with MRT processing is only preferred in environments having a high signal-to-noise ratio, low mobility, and small Ricean $K$ -factors. [ABSTRACT FROM AUTHOR]

This paper considers a multicell downlink (DL) massive MIMO system operating over Ricean fading channels in which each base station (BS) is equipped with a massive antenna array, while each user has a single antenna. We explore equal-gain transmission (EGT), line-of-sight (LOS) component-based EGT (LOS-EGT) and maximum-ratio transmission (MRT) under imperfect channel state information. Closed-form expressions for lower bounds of the achievable rates are derived for EGT over Ricean and Rayleigh fading channels, and for LOS-EGT and MRT over Ricean fading channels. With the obtained closed-form expressions, various power scaling laws concerning DL data transmit power and uplink (UL) pilot transmit power are established and discussed. In particular, it is found that, as the number of BS antennas $M$ grows unlimited, the lower bounds on the rates achieved with EGT, LOS-EGT and MRT schemes approach infinity and are not affected by pilot contamination, while the DL data transmit power and UL pilot transmit power can be scaled down proportionally to ${M^{ -a}}$ and ${M^{ - b}}$ (where $0 \leq a < 1$ and $b > 0$), respectively. Numerical results corroborate the tightness and accuracy of these closed-form expressions and they also show that, when the number of antennas and intercell interference level are large, compared to the MRT, EGT and LOS-EGT are more resistant to intercell interference and pilot contamination. [ABSTRACT FROM AUTHOR]

This paper considers the downlink transmission in a single-cell proximity co-located massive MIMO (PC-MIMO) system, in which antennas of the base station (BS) are divided into multiple antenna units (AUs) and these AUs are connected to a backhaul processor unit (BPU). The AUs are geographically deployed very close to the BPU which controls them to serve all users in the cell. Considering correlated Rician fading channels, we derive a closed-form expression of the sum spectral efficiency under the maximum-ratio transmission and statistical channel state information at each user. Based on the obtained expression, it is shown that as the number of AUs approaches infinity, the sum spectral efficiency approaches a constant value. Furthermore, we also give a closed-form expression of the average sum spectral efficiency when the arrival angles to the users seen from the AUs are varying. For the cases of correlated Rayleigh fading and uncorrelated Rician fading, we show how to find the number of antennas at each AU to maximize the average sum spectral efficiency when the total number of antennas at the BS is fixed. [ABSTRACT FROM AUTHOR]