Massive MIMO under Double Scattering Channels: Power Minimization and Congestion Controls

Abstract

This paper considers a massive MIMO system under the double scattering channels. We derive a closedform expression of the uplink ergodic spectral efficiency (SE) by exploiting the maximum-ratio combining technique with imperfect channel state information. We then formulate and solve a total uplink data power optimization problem that aims at simultaneously satisfying the required SEs from all the users with limited power resources. We further propose algorithms to cope with the congestion issue appearing when at least one user is served by lower SE than requested. Numerical results illustrate the effectiveness of our proposed power optimization. More importantly, our proposed congestion-handling algorithms can guarantee the required SEs to many users under congestion, even when the SE requirement is high.

Sum-Rate Analysis of a Multi-Cell Multi-User MISO System Under Double Scattering Channels

“…Both works aimed at finding closed-form expressions of the sum-rate of the uplink multi-cell massive MIMO system under a double scattering channel model and optimizing the power allocation accordingly. However, the works in [28] and [29] differ from our work as follows. First, the downlink link was not studied and the effect of the number of scatterers, the number of antennas, and the number of users was not well studied under different regimes to understand their role on the performance.…”

“…However, the analysis in this work was purely empirical; no closed-form expressions were derived and the only insights were drawn through simulations. Additionally, there are two other unpublished works related to the double scattering channels [28], [29]. Both works aimed at finding closed-form expressions of the sum-rate of the uplink multi-cell massive MIMO system under a double scattering channel model and optimizing the power allocation accordingly.…”

“…Second, a strong orthogonality requirement for the users' channel correlation matrices was considered in [28], [29], which does not hold generally in practice.…”