雷达通信一体化数字预失真研究

发布时间：2018-12-19 13:10

【摘要】：雷达作为航空航天和军事领域的重要电子设备,传统的先侦察后处理方式以及低速实时数据处理能力越来越不能够满足要求。随着科技的发展和进步,多终端,小型化,多功能的要求使得雷达通信一体化系统的核心技术研究非常迫切。雷达通信一体化系统中,雷达功率放大器对普通数字调制信号的非线性影响是制约一体化系统高速通信的主要障碍,本文围绕这个问题,主要研究了雷达功率放大器的非线性特性,对数字调制信号产生的影响,以及基于多项式和查表法的不同数字预失真技术等。雷达系统的功率放大器通常工作在深度饱和区以获得较高的发射功率,致使信号产生严重的失真,造成接收端误码率的升高甚至无法解调信号,文中仿真了不同功放模型和雷达功放的特性曲线,证明雷达功率放大器具有记忆效应,无记忆非线性系统模型对雷达功放的非线性描述误差较大,记忆多项式功放模型能够精确描述雷达功放的非线性特性。之后本文仿真分析了雷达功放模型对三种调制方式QPSK,pi/4QPSK和16QAM信号的失真影响,仿真结果表明三种调制方式中,16QAM信号受对功放非线性效应最敏感,该系统的数字预失真算法必须首先能够成功校正16QAM信号。雷达通信一体化系统的前提是在原有雷达设备基础上增添通信信号处理模块,不能影响雷达本身工作,因此本文对较为具有工程实践意义的无记忆多项式预失真算法,Hammerstein预失真算法,记忆多项式算法以及查表法做了主要研究,对预失真的自适应算法做了仿真性能分析,在LMS算法的基础上作出改进以更易于硬件的实现。仿真分析了不同算法对16QAM信号通过雷达功放模型的星座图,眼图以及误码率的改善效果,仿真结果表明无记忆预失真模型对无记忆功放改善效果显著,但无法有效补偿记忆功放模型非线性效应。Hammerstein预失真模型对Wiener功放模型的非线性效应和线性记忆效应线性化效果良好,但无法描述雷达功放的非线性记忆效应。记忆多项式预失真模型能够对雷达功放模型起到良好的线性化效果,阶数和记忆深度越高,补偿效果越好,同时实现难度越大。为了降低记忆多项式的实现难度,将记忆多项式模型与查表法相结合,使用查找表存储多项式系数的方式构建模型,仿真结果表明,该模型对雷达功放的线性化效果显著,星座图和眼图得到改善,误码率接近线性功放情况。
[Abstract]:Radar as an important electronic equipment in aerospace and military fields, the traditional methods of reconnaissance and processing, as well as the ability of low-speed real-time data processing, are more and more unable to meet the requirements. With the development and progress of science and technology, the requirements of multi-terminal, miniaturization and multi-function make the research of the core technology of radar communication integration system very urgent. In the integrated radar communication system, the nonlinear influence of radar power amplifier on the common digital modulation signal is the main obstacle restricting the high-speed communication of the integrated system. The nonlinear characteristics of radar power amplifier, the influence on the digital modulation signal, and the different digital predistortion techniques based on polynomial and table searching method are studied in this paper. The power amplifier of radar system usually works in the depth saturation region to obtain high transmission power, which causes serious distortion of the signal, resulting in the increase of the bit error rate at the receiving end and even the failure to demodulate the signal. In this paper, different power amplifier models and characteristic curves of radar power amplifier are simulated. It is proved that radar power amplifier has memory effect, and the nonlinear description error of radar power amplifier by the model of memoryless nonlinear system is large. The memory polynomial power amplifier model can accurately describe the nonlinear characteristics of radar power amplifier. The simulation results show that the 16QAM signal is most sensitive to the nonlinear effect of the three modulation modes, and the simulation results show that the 16QAM signal is most sensitive to the nonlinear effect of the power amplifier, and the simulation results show that the 16QAM signal is most sensitive to the nonlinear effect of the amplifier. The digital predistortion algorithm of the system must be able to successfully correct the 16QAM signal. The premise of the integrated radar communication system is to add the communication signal processing module to the original radar equipment, which can not affect the work of the radar itself. Therefore, this paper deals with the memoryless polynomial predistortion algorithm, which has practical significance in engineering. The Hammerstein predistortion algorithm, memory polynomial algorithm and table lookup method are mainly studied. The simulation performance of the adaptive predistortion algorithm is analyzed. Based on the LMS algorithm, the improvement is made to facilitate the hardware implementation. The effects of different algorithms to improve the constellation diagram, eye diagram and bit error rate of 16QAM signal through radar power amplifier model are analyzed. The simulation results show that the memoryless predistortion model can improve the memoryless power amplifier significantly. But the nonlinear effect of memory amplifier model can not be compensated effectively. Hammerstein predistortion model has good linearization effect on Wiener power amplifier model, but it can not describe the nonlinear memory effect of radar power amplifier model. The memory polynomial predistortion model can give a good linearization effect to the radar power amplifier model. The higher the order and memory depth, the better the compensation effect and the more difficult the realization. In order to reduce the difficulty of realizing the memory polynomial, the memory polynomial model is combined with the look-up table method, and the model is constructed by using the look-up table to store the polynomial coefficients. The simulation results show that the model has a remarkable linearization effect on the radar power amplifier. Constellation diagram and eye chart are improved, error rate is close to linear power amplifier.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN958

【相似文献】