记忆非线性功率放大器的高效预失真

记忆非线性放大器的预失真问题一直是预失真技术的难点。通常采用Volterra级数、Hammerstein模型和神经网络等模型的记忆预失真都存在形式复杂、自适应困难的缺点。文章通过增加两个延时环节将基于多项式的无记忆放大器的高效预失真结构推广到有记忆放大器的预失真中,并联合一种简单的带抽头延时的非线性多项式模型作为记忆预失真器模型实现了记忆非线性放大器的快速、高效的线性化。仿真结果表明,利用所提出的预失真方案能快速实现记忆放大器的预失真,而且显著提高了线性化性能。

High-efficient adaptive predistortion scheme for nonlinear power amplifier with memory

Predistortion for nonlinear amplifier with memory always is very difficult.Traditional method using Voherra,Hammerstein and neural network model is not only complex,but their adaptation also not easy.High-efficient predistortion structure based on polynomial in memoryless predistortion is extended to memory predistortion by adding two-delay lines,and a simple polynomial model with tapped delay linear is used as the memory predistorter model.Combined the extended structure and the simple memory predistorter model can realize the linearization of the amplifier with memory.Simulation results demonstrated that the proposed scheme posses faster convergence and better linearization performance.