Low-Dimensional Post-Distortion for Nonlinear Power Amplifier in MIMO Systems

To satisfy the growing demand for high throughput, mobile communication systems have been evolving towards more antennas, and the array gain and spatial multiplexing provided by multiple input multiple output (MIMO) have been widely used to improve the capacity. However, in practice, the MIMO system inevitably contains nonlinearity, due to the power amplifiers (PAs) functioning in nonlinear regions. In addition, because of the dynamic trapping, dynamic self-heating and other factors, PA will present long-term memory effect, which intensifies the nonlinear behavior. The strong nonlinearity will distort the transmitted signals, resulting in deterioration of communication quality. To alleviate this negative impact, a memory polynomial with memory elimination (MP-ME) scheme is introduced, where the nonlinear and memory distortions are separated apart and processed independently, thus avoiding the calculation of high-dimensional matrix. Further, a memory characteristic extraction algorithm is presented in this paper, which extends the MP-ME scheme to strong complex nonlinear systems. Simulation results show that MP-ME can achieve excellent linearization performance, making itself a promising post-distortion solution.