基于测量的记忆多项式功率放大器特征模型

Measurement-based memory polynomial behavioral modeling of RF power amplifiers

给出了基于ADS-ESG-VSA89600测量的功率放大器记忆多项式特征模型,并使用复值Givens旋转基础的QR分解递归最小二乘算法进行模型系数更新。通过使用安捷伦连接结论测试平台ADS-ESG-VSA89600去收集实际功率放大器输入和输出的时域包络数据来进行模型的生成和验证。与批量和递归最小二乘算法相比,QR分解基础的递归最小二乘算法具有良好的数值鲁棒性和规则的结构,因此能够明显的提高模型的精确性并适合用于硬件实现。仿真和测量结果表明,本模型能够精确的复制出功率放大器的宽带动态特性,并真实的捕捉到放大器的非线性和记忆效应。

A memory polynomial approach for dynamic behavioral modeling of nonlinear power amplifier with memory based on the ADS-ESG-VSA89600 measurement is proposed in the paper. The proposed approach utilizes the complex QR-decomposition based recursive least squares （QRD-RLS） algorithm, which is implemented using the com- plex Givens rotations to update the coefficient vector of the memory polynomial modeling. By using the Agilent con- nected-solution test bench ADS-ESG-VSA89600, the complex envelope data from the measured input and output of the real PA is captured, and which is used to extract and validate the behavioral model proposed. Comparing with the stan- dard least squares algorithms, in batch and recursive process, the QRD-RLS algorithm has the characteristics of good numerical robustness and regular structure, and can obviously improve the memory polynomial modeling accuracy and well suite for hardware implementation. The results of simulation and experimentation show that the model can repro- duces the dynamical behavior of the amplifier accurately, and capture the nonlinear and memory effects of the amplifier with high fidelity.