Large-Scale-Fading Decoding in Cellular Massive MIMO Systems With Spatially Correlated Channels

Abstract

Massive multiple-input-multiple-output (MIMO) systems can suffer from coherent intercell interference due to the phenomenon of pilot contamination. This paper investigates a two-layer decoding method that mitigates both coherent and non-coherent interference in multi-cell Massive MIMO. To this end, each base station (BS) first estimates the channels to intracell users using either minimum mean-squared error (MMSE) or element-wise MMSE (EW-MMSE) estimation based on uplink pilots. The estimates are used for local decoding on each BS followed by a second decoding layer where the BSs cooperate to mitigate inter-cell interference. An uplink achievable spectral efficiency (SE) expression is computed for arbitrary two-layer decoding schemes. A closed-form expression is then obtained for correlated Rayleigh fading, maximum-ratio combining, and the proposed large-scale fading decoding (LSFD) in the second layer. We also formulate a sum SE maximization problem with both the data power and LSFD vectors as optimization variables. Since this is an NP-hard problem, we develop a low-complexity algorithm based on the weighted MMSE approach to obtain a local optimum. The numerical results show that both data power control and LSFD improve the sum SE performance over singlelayer decoding multi-cell Massive MIMO systems.

Large-Scale Fading Precoding for Maximizing the Product of SINRs

“…This prior work considered spatially uncorrelated Rayleigh fading and the LSFP and LSFD weights are optimized using the max-min fairness criterion. Then, [2] and [11] derived the SE with LSFD in the uplink for spatially correlated Rayleigh fading channels. There are no comparable studies for LSFP.…”

“…Each cell is composed of an M -antenna base station (BS) and K single-antenna users. We assume all BSs are connected to a central network controller in accordance with the existing literature [2], [9]. The conventional block-fading model [4] is assumed where the channel between each BS antenna and user is a complex static scalar in one coherence block of τc channel uses and take independent realization in each block.…”

“…Note that the vectors {ḡ r lk } and covariance matrices {R r lk } describe the long-term channel effects and change more slowly compared to small-scale fading characteristics. We assume that all BSs have the knowledge of {ḡ r lk } and covariance matrices R r lk in accordance with the multi-cell massive MIMO literature [2], [12]. However compared to most of the existing literature, we consider a more realistic scenario in which phase of the LOS component varies at the same pace as the small-scale fading, due movement, and the phase is unknown.…”

“…Massive MIMO (multiple-input multiple-output) is one of the key technologies of 5G cellular systems and commercial deployments started in 2018 [1]. Massive MIMO provides high spectral efficiency (SE) to a large number of users using the same time-frequency resources by an efficient spatial multiplexing and large number of antennas [2][3][4][5][6][7]. In [8], it is shown that the SE of the users is limited by pilot contamination which is the result of interference from the other cells' users whose pilot signals are not orthogonal to the considered cell's users.…”

“…One trivial remedy is to increase the length of the pilot signals to make them all orthogonal throughout the network. However, due to limited coherence block length in practical communication systems, this approach is not efficient [2], [9]. Another solution to pilot contamination is to exploit the spatial correlation among the BS antennas [10].…”