卫星数传发射机线性化技术研究

发布时间：2018-01-26 06:38

本文关键词： 卫星通信数字预失真查找表自适应算法 FPGA　出处：《中国科学院大学(中国科学院国家空间科学中心)》2017年硕士论文　论文类型：学位论文

【摘要】：在卫星通信中,为了更高效的利用有限的频谱资源,一系列频谱利用率较高的调制方式,如M-QAM(Multi Quadrature Amplitude Modulation)、M-APSK(Multi Amplitude Phase Shift Keying)等正逐渐被应用于数据传输系统。常用的高阶调制信号大多为非恒定包络调制信号,与恒定包络信号相比其峰均比较高。由于卫星通信系统特殊的工作环境,功率放大器需要工作在饱和点或邻近饱和点。当非恒定包络信号经过饱和功放时,将产生严重的非线性失真,非线性失真会使输出信号的带内与带外产生严重失真。为了补偿功率放大器对信号产生的失真,本文针对数字预失真技术进行研究,并对相关算法进行了仿真、改进和验证工作。数字预失真线性化技术主要包括查找表技术和自适应数字预失真技术。针对卫星通信数据传输系统的应用需求,本文对数字预失真技术在卫星数传系统中的应用展开研究。主要研究包括以下几个方面:⒈针对无记忆功放模型和记忆功放模型,采用查找表及自适应预失真技术进行数字预失真仿真,并分析自适应算法参数对预失真效果的影响。⒉基于最小二乘算法和记忆多项式模型,提出了一种实数型QR分解LS预失真算法。该自适应预失真算法对传统复数型LS算法进行改进,避免了复数矩阵QR分解的复杂运算以及传统LS算法矩阵求逆运算,实现了利用实数矩阵精确快速求解预失真器参数。⒊对实数型QR分解LS算法进行理论推导证明,并对该算法进行Matlab仿真验证。在Matlab中搭建仿真环境,采用16QAM调制信号与记忆多项式功放记忆模型,并与传统LS算法进行对比分析。⒋搭建基于矢量信号发生器和频谱仪的半实物仿真平台,对本文提出的实数型QR分解LS算法进行半实物仿真测试,验证了算法的正确性和有效性。⒌对本文所提出的数字预失真系统核心算法进行FPGA代码设计,并进行仿真,包括:数据流幅度差相关法环路延时算法、基于脉动阵列的实数型QR分解LS算法。
[Abstract]:In satellite communications, in order to make more efficient use of limited spectrum resources, a series of spectrum efficient modulation methods. For example, M-QAMN Multi Quadrature Amplitude Modulation. M-APSKK Multi Amplitude Phase Shift Keying. Equal-order modulation is gradually being used in data transmission systems. Most of the high-order modulation signals are unsteady envelope modulation signals. Because of the special working environment of the satellite communication system, the power amplifier needs to work at or near the saturation point. When the unsteady envelope signal passes through the saturation power amplifier. In order to compensate for the distortion caused by power amplifier, digital predistortion technology is studied in this paper. The digital predistortion linearization technology mainly includes look-up table technology and adaptive digital predistortion technology, aiming at the application requirements of satellite communication data transmission system. In this paper, the application of digital predistortion technology in satellite data transmission system is studied. The main research includes the following aspects: 1 for memoryless power amplifier model and memory power amplifier model. The digital predistortion simulation is carried out by using lookup table and adaptive predistortion technique. The effect of adaptive algorithm parameters on predistortion effect is analyzed. 2. The least square algorithm and memory polynomial model are used to analyze the effect of the adaptive algorithm parameters on the predistortion effect. A real QR decomposition LS predistortion algorithm is proposed, which improves the traditional complex LS algorithm. The complex operation of QR decomposition of complex matrix and the matrix inversion of traditional LS algorithm are avoided. The real QR decomposition LS algorithm is theoretically deduced and proved by using real matrix to solve the predistorter parameter accurately and quickly. The simulation environment is built in Matlab, and the memory model of 16QAM modulation signal and memory polynomial power amplifier is adopted. And compared with the traditional LS algorithm, .4 build a hardware-in-the-loop simulation platform based on vector signal generator and spectrometer, and carry out the hardware-in-the-loop simulation test of the real QR decomposition LS algorithm proposed in this paper. The correctness and validity of the algorithm is verified. 5. The core algorithm of digital predistortion system proposed in this paper is designed by FPGA code, and simulated, including: data flow amplitude difference correlation loop delay algorithm. Real QR decomposition LS algorithm based on pulsating Array.
【学位授予单位】：中国科学院大学(中国科学院国家空间科学中心)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN722.75;TN927.2

【相似文献】