基于FPGA的5G功放DPD模型设计与实现

    Design and Implementation of 5G Power Amplifier Digital Predistortion Model Based on FPGA

    • 摘要: 为了能够有效抑制5G射频功放的强非线性,文中设计并实现了基于现场可编程门阵列多处理器片上系统(FPGA MPSOC)平台的简化广义幅值选择仿射(SGMSA)数字预失真器,并使用并行结构提高带宽处理能力和处理速度。在同一架构上实现了修正的幅值选择性仿射(MMSA)模型和广义记忆多项式(GMP)模型并将其与本文模型进行对比,在硬件资源占用方面SGMSA资源占用率更低,其查找表资源分别减少1%和6%,触发器资源减少1%,计算单元资源分别减少14%和46%,片上总功耗分别减少0.137 W和0.216 W。采用100 MHz 5GNR信号通过FPGA平台在2.6 G和3.5 G功放上进行验证,SGMSA上边带和下边带临信道功率比性能明显优于MMSA模型与GMP模型以及其他论文同功放下模型。文中在FPGA实现了SGMSA模型电路,证明其具有较好的预失真性能以及实际工程应用价值。

       

      Abstract: In order to effectively suppress the strong nonlinearity of 5G RF power amplifiers, this paper designs and implements a SGMSA digital predistorter based on the FPGA MPSOC platform, and uses a parallel structure to improve bandwidth processing capability and processing speed. The MMSA model and GMP model are implemented on the same architecture and compared with the model in this paper. In terms of hardware resource occupation, the SGMSA resource occupation rate is lower, with a reduction of 1% and 6% in lookup table resources, a reduction of 1% in trigger resources, a reduction of 14% and 46% in computation unit resources, and a reduction of 0.137 W and 0.216 W in total on-chip power consumption, respectively. Using a 100 MHz 5GNR signal, the performance of the SGMSA upper and lower sideband near-channel power ratio is significantly better than the MMSA model and GMP model, as well as other papers with the same power amplifier. This paper implements the SGMSA model circuit on FPGA, proving its good predistortion performance and practical engineering application value.

       

    /

    返回文章
    返回