基于分离法和分段线性模型的高速预失真技术

Memory piecewise linear predistorter of the HPA with memory based on the separation method

本文提出一种基于分离法和分段线性模型的有记忆HPA(High Power Amplifier)预失真技术,并给出基于查找表的FPGA高速实现方案。该预失真方案首先应用进化算法自适应辨识出有记忆HPA的(LTI+分段线性函数)模型,然后根据分离法原理,求出预失真器模型的参数。本文的分离法和文献的方法相比,运算量较小,而且精度更高,易于高速FPGA实现,适合于高速卫星发射机。matlab仿真结果证明该方法的性能很好。

This paper puts forward the memory piecewise linear predistorter based on the separation method and bring out a FPGA implementation scheme based on the look-up table with low complexity when used in the high-speed predistortion system. After approximating the nonlinear static part by a piecewise linear function, its parameters are estimated by using the evolutionary computation approach,while the linear system part is estimated by the least squares method. The scheme identifies the HPA model,then gets the model parameters of the predistorter according to the separation method. Compared with the method in reference[ 12 ] , the scheme has much less computation but higher precision. The adaptive predistortion technique is easy to implement and fit for the satellite transmitter. It is proved to be correct through Matlab simulation.