On the Impact of Hardware Impairments in Noncoherent Massive MIMO Systems

Massive MIMO systems, where a very large number of antennas are employed at the base station, are one of the key technologies to increase the capacity in multi-user MIMO scenarios. This comes at the price of raised hardware costs and power consumption, which scales linearly with the number of receiver chains. Hence, a low-cost and power-efficient hardware design is of special interest. However, such a low-complexity radio frontend impairs the quality of the received signal, impacting the overall system performance. In this paper, the robustness to hardware impairments at the base station is analyzed on a theoretical basis when noncoherent detection schemes are applied in the uplink transmission. The effects on the system performance are assessed by means of numerical simulations. Behavioral and stochastic impairment models are utilized to cover the influence of the main hardware components, i.e., amplifiers, local oscillators, and analog-to-digital converters. A cluster-based channel model adapted to the massive MIMO setting forms the basis for the evaluation. It is shown that highly-linear amplifiers are integral for optimum performance. However, low-resolution ADCs and oscillators of moderate quality can be employed without sacrificing performance, saving power and costs.