卫星系统高数据吞吐量均衡器的算法及结构研究

发布时间：2018-05-19 05:20

本文选题：卫星数据传输 + 均衡算法　；参考：《燕山大学》2014年博士论文

【摘要】：随着通信网络中信息交换量的加大，现代卫星高速数据传输系统的均衡器，需在数据吞吐量、幅度误差和相位误差抑制能力、算法收敛速度和稳态误差等方面具有更高的性能指标。当前Bussgang (巴斯冈)类算法，，由于算法性能显著和计算复杂度低，使其在理论研究和实际工程中呈现巨大价值。但受限于Bussgang类算法本身固有机制约束，使其对圆包络星座图特征的调制信号在相位误差抑制上，对多模值高阶调制信号在均衡性能上，以及对算法工程实现中传输数据的吞吐量上，均呈现出不足。因此，本文以系统群时延建模的非线性相频特性卫星信道为基础，针对均衡算法的相位误差抑制能力、多模值调制信号的算法性能和均衡器的数据吞吐量三个问题展开研究。针对圆包络单模值和多模值星座图特征的调制信号，分别提出星座图相位匹配算法和象限恒定相位算法，以实现对传输信号相位误差的抑制。根据圆包络单模值调制信号的星座图相位特征，构建由正弦函数组成的星座图相位误差匹配项，并通过与恒模算法代价函数进行线性加权，进而得到星座图相位匹配算法。根据圆包络多模值星座图相位特征，依据各象限相位统计量为常数值的假设，构建相位误差匹配项，进而得到象限恒定相位算法。针对多模值圆包络和矩形包络星座图特征调制信号，分别提出双模式恒模均衡算法和双模式多模均衡算法，以使算法在具有稳定收敛性能同时，实现算法与多模值调制信号的星座图特征相匹配。算法通过递归方程估计均衡器瞬时收敛状态，根据均衡器收敛过程中输出信号分布特征的变化规律，得到收敛状态的判决阈值，进而实现对均衡器的双模式控制。在算法建立之初，均衡器工作于散度常量模式，以使均衡器获得稳定的收敛性能；在算法建立之后，均衡器工作于散度变量模式，以通过算法与星座图匹配，实现均衡器收敛速度的提高和稳态误差的降低。构建并行滤波器和流水线算法迭代结构均衡器，以消除电子器件对均衡器数据吞吐量的约束，提高均衡器数据吞吐量；提出适应该结构均衡算法，以使均衡器在数据吞吐量提高的同时，改善均衡器整体性能。该结构中，并行滤波器用于消除电子器件最大运行速率对数据吞吐量的约束，流水线结构用于降低算法迭代过程对电子器件实时性的削弱，但该结构同时增大算法迭代周期，降低算法对传输信号的利用率。因此，根据梯度下降原理和均衡器抽头系数迭代过程的变化规律，提出多梯度模式均衡算法。综合全文研究内容，针对各种包络和模值星座图特征中代表性调制方式，正交相移键控、8阶相移键控、16阶正交幅度调制和16阶幅相键控调制信号，给出具有消除电子器件对数据吞吐量约束，实现相位误差抑制和多模值特征相匹配能力的均衡器设计方案。搭建基于群时延特征的卫星数据传输测试平台，使用电子器件FPGA实现各算法设计方案，并通过FPGA输出信号的误差值和星座图的实验结果分析，验证算法有效性。
[Abstract]:With the increase of the information exchange in the communication network , the equalizer of the high - speed data transmission system of the modern satellite has higher performance index in terms of data throughput , amplitude error and phase error suppression capability , algorithm convergence speed and steady - state error .

A constellation diagram phase matching algorithm and a quadrant constant phase algorithm are respectively proposed for the modulation signals of the round envelope single - mode value and the multi - mode value constellation diagram , so as to realize the suppression of the phase error of the transmission signal . According to the constellation phase characteristics of the circular envelope single - mode value modulation signal , a constellation phase matching algorithm is constructed . According to the phase characteristics of the circular envelope multi - mode value constellation map , a phase error matching term is constructed according to the assumption that the phase statistic of each quadrant is a constant value , and then a quadrant constant phase algorithm is obtained .

A dual - mode constant - mode equalization algorithm and a dual - mode multi - mode equalization algorithm are proposed for the multi - mode value circle envelope and the rectangular envelope constellation feature modulation signal . The algorithm is matched with the constellation diagram of the multi - mode value modulated signal . The algorithm estimates the instantaneous convergence state of the equalizer through the recursive equation , and then obtains the decision threshold of the convergence state according to the variation law of the output signal distribution characteristics in the convergence process of the equalizer .
After the algorithm is established , the equalizer operates in the divergence variable mode to match the constellation diagram with the algorithm to realize the improvement of the convergence speed of the equalizer and the reduction of the steady state error .

constructing an iterative structure equalizer of the parallel filter and the pipeline algorithm so as to eliminate the constraint of the electronic device on the equalizer data throughput and improve the data throughput of the equalizer ;
in that structure , the parallel filter is used for eliminating the constraint of the maximum operating speed of the electronic device on the data throughput , and the streamline structure is used for reducing the effect of the algorithm iteration process on the real - time performance of the electronic device , but the structure simultaneously increases the iteration period of the algorithm and reduces the utilization rate of the transmission signal .

In this paper , an equalizer design scheme is given , which can eliminate the constraint of electronic devices on data throughput , realize phase error suppression and multi - mode value characteristic matching .
【学位授予单位】：燕山大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN927.2;TN715

【参考文献】