短时突发信号的均衡技术研究

发布时间：2018-12-18 17:26

【摘要】：短时突发通信因其发送信号的短时性和突发性而具有较强的抗干扰与抗截获能力,且目前已有多方面的应用,诸如潜艇通信和北约国家目前大量装备的Link16数据链等。短时突发信号(一般只有几十或几百个符号)对于侦察方而言侦察或干扰难度较大,而合作方在接收信号时难度也要大于一般的长时间连续信号。为了减小码间干扰,降低误码率,保证通信质量,需要在接收端对短时突发信号进行信道均衡。本文将针对短时突发信号进行自适应均衡研究。本文主要的工作以及研究成果如下:(1)简单介绍短时突发信号及其主要特点(短时性和突发性),根据其信号特征分析期望的收敛性能,并引出“数据重用”的自适应均衡思想。(2)简单介绍“数据重用”思想以及基于该思想的代表算法,即仿射投影算法和仿射投影符号算法,阐述了其产生及应用背景,并从理论上给出其约束最优化模型,分析主要参数(迭代步长、投影阶数)对算法性能的影响。(3)针对仿射投影算法,本文从可变迭代步长、可变投影阶数和组合算法三个角度进行分析。对于可变迭代步长,文中利用矩阵奇异值分解以及伪逆的知识,对常规的仿射投影算法的迭代公式进行修改,然后从均方偏差的角度分析了该修正算法,并推导出可变步长的迭代公式。对于可变投影阶数,文中推导出了仿射投影算法的均方偏差的近似表达式,令均方偏差的增量最小化以确定每次迭代的可变投影阶数。针对组合算法,文中仿照现有的组合算法的思路,自适应组合可变步长的仿射投影算法和归一化最小均方算法,并从均方偏差的角度分析了用于组合的混合参数。(4)针对仿射投影符号算法,根据“误差向量比其符号信息包含更多的关于未知信道的信息”,文中首先对算法的迭代公式进行修改,然后从后验估计误差的角度分析该修正算法,并推导出了修正算法的可变步长。(5)对于文中研究的新算法,本文均从理论推导以及计算机仿真的角度分析了不同情况下的收敛性能、不同参数对算法性能的影响,以及处理短时突发信号时,应该如何设置这些参数。(6)对本文工作进行了总结与反思,并对后续研究方向进行了分析与展望。
[Abstract]:Short-time burst communication has strong anti-jamming and anti-interception capability because of its short-time and sudden signal transmission, and has been used in many fields, such as submarine communication and Link16 data link, which is a large number of equipment in NATO countries, and so on. Short-time burst signals (usually only a few dozen or hundreds of symbols) are more difficult to detect or interfere with for the reconnaissance side, while the cooperative side is also more difficult than the conventional long-term continuous signal in receiving the signal. In order to reduce inter-symbol interference (ISI), reduce bit error rate (BER) and ensure communication quality, it is necessary to equalize short-time burst signals at the receiving end. In this paper, adaptive equalization of short-time burst signals is studied. The main work and results of this paper are as follows: (1) briefly introduce the short-time burst signal and its main characteristics (short-time and sudden), and analyze the desired convergence performance according to its signal characteristics. The idea of adaptive equalization of "data reuse" is introduced. (2) the idea of "data reuse" and its representative algorithms, namely affine projection algorithm and affine projection symbol algorithm, are briefly introduced. The constrained optimization model is given theoretically, and the influence of main parameters (iterative step size, projection order) on the performance of the algorithm is analyzed. (3) for affine projection algorithm, the variable iterative step size is used in this paper. Variable projection order and combination algorithm are analyzed from three angles. For the variable iterative step size, the iterative formula of the conventional affine projection algorithm is modified by using the knowledge of matrix singular value decomposition and pseudo inverse, and the modified algorithm is analyzed from the angle of mean square deviation. An iterative formula of variable step size is derived. For variable projection order, the approximate expression of mean square deviation of affine projection algorithm is derived, and the increment of mean square deviation is minimized to determine the variable projection order of each iteration. For the combination algorithm, this paper imitates the existing combination algorithm, adaptively combines the variable step size affine projection algorithm and the normalized least mean square algorithm. The mixed parameters used for combination are analyzed from the angle of mean square deviation. (4) for affine projection symbol algorithm, according to "error vector contains more information about unknown channel than its symbol information", This paper first modifies the iterative formula of the algorithm, then analyzes the modified algorithm from the point of view of the posteriori estimation error, and deduces the variable step size of the modified algorithm. (5) for the new algorithm studied in this paper, In this paper, the convergence performance under different conditions, the influence of different parameters on the performance of the algorithm, and the processing of short time burst signals are analyzed from the angle of theoretical derivation and computer simulation. How to set these parameters. (6) summarize and reflect on the work of this paper, and analyze and prospect the future research direction.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN911.5

【参考文献】