Multi-Rate Hybrid Predistortion with Reduced Sampling Rate and Resolution for 6G Power Amplifier Linearization

abstract-To achieve cost-effective and energy-efficient 6G base station (BS) deployment, digital predistortion (DPD) is essential to concurrently maintain high power efficiency and optimal linearity of power amplifiers (PA). However, dramatically increased communication signal bandwidth and carrier frequency in 6 G lead to extremely high sampling rate requirements in conventional DPD, which significantly increases DPD implementation cost and overall power consumption. In this paper, we propose a novel multi-rate hybrid predistortion scheme to reduce the cost and power consumption for 6 G PA linearization by lowering the sampling rate and analog-to-digital converter (ADC) resolution concurrently. Specifically, a new adaptive PA distortion estimation method is proposed using pseudo-preambles with varying lengths to mitigate the impact of low ADC resolution and low feedback loop sampling rate, while guaranteeing the estimation accuracy without inducing additional hardware complexity. Furthermore, a new PA distortion characteristics specified multi-rate hybrid predistortion scheme is proposed that tailors different sampling rates to the decomposition and linearization of estimated PA static and dynamic distortions, thus further mitigating low baseband sampling rate impact with guaranteed linearization performance. Simulation results demonstrate the proposed predistortion scheme achieves satisfactory PA linearization performance and power efficiency with dramatically reduced sampling rates and resolution.