Abstract
This paper addresses multi-user multi-cluster massive
multiple-input-multiple-output (MIMO) systems with non-orthogonal
multiple access (NOMA). Assuming the downlink mode, and taking into
consideration the impact of imperfect successive interference
cancellation (SIC), an in-depth analytical analysis is carried out, in
which closed-form expressions for the outage probability and ergodic
rates are derived. Subsequently, the power allocation coefficients of
users within each sub-group are optimized to maximize fairness. The
considered power optimization is simplified to a convex problem, which
makes it possible to obtain the optimal solution via Karush-Kuhn-Tucker
(KKT) conditions. Based on the achieved solution, we propose an
iterative algorithm to provide fairness also among different sub-groups.
Simulation results alongside with insightful discussions are provided to
investigate the impact of imperfect SIC and demonstrate the fairness
superiority of the proposed dynamic power allocation policies. For
example, our results show that if the residual error propagation levels
are high, the employment of orthogonal multiple access (OMA) is always
preferable than NOMA. It is also shown that the proposed power
allocation outperforms conventional massive MIMO-NOMA setups operating
with fixed power allocation strategies in terms of outage probability.