PDM相干检测系统均衡算法的研究

发布时间：2018-06-20 21:38

本文选题：波分复用 + 偏振复用　；参考：《南京邮电大学》2017年硕士论文

【摘要】：随着各种数据量的爆炸式增长和带宽限制,基于光纤放大器和传统密集波分复用的光通信系统已渐渐满足不了快速增长的信息需求。随着近几年对偏振复用技术的研究,采用两个正交偏振态独立传输信息,其不需要占用额外带宽,使得在已有复用技术的基础上增加了传输容量,且进一步提高了频谱利用率。与相干检测技术相结合,被广泛的应用在现代高速光通信系统中。在光通信系统中,随着光纤放大器的应用,光纤损耗对通信质量的影响可以忽略,所以目前人们把研究重点放在了色散补偿上。目前比较成熟的色散补偿技术主要在光域中进行的,多针对于直接检测系统,已经不能满足高速相干光通信的要求。本文将针对数字处理模块所使用的自适应算法进行研究,解决算法收敛速度和收敛精度对立的问题,并对高阶调制方式中使用的算法进行了深入的研究。文章的具体研究内容如下:(1)本文对电域均衡算法做了大量的理论分析,推导出算法收敛速度和收敛时间对立的问题,并介绍了目前针对这个问题的两种优化方案,通过对改进方案的分析发现其计算复杂要高于固定步长算法。所以本文针对这一问题提出了分段迭代步长的改进算法,分段步长中包含固定步长和变步长两种方式,将这种改进方式应用于最小二乘法和恒模算法中。通过偏振复用QPSK调制系统验证了方案的可行性。结果表明改进的分段步长迭代算法,在收敛速度和精度上要好于一般改进算法,且算法的计算复杂度较一般改进算法低。(2)对于16QAM调制方式方式,为了保证系统的高频谱利用率,目前所使用的均衡算法为无需参考序列的恒模算法。但是16QAM调制方式模值的个数不唯一,一般的恒模算法不能直接使用,针对这一问题提出了一种基于虚实分开计算的改进算法,这种算法可以降低模值个数,但是不能将模值变为唯一,且算法计算复杂度较高。所以针对这些问题本文提出了基于判决理论的改进恒模算法,这种算法利用接收信号模值唯一的特点实现恒模算法。通过分析,基于判决理论算法与虚实处理算法相比,结构更简单、均衡效果更好,计算复杂度更低。
[Abstract]:With the explosive growth of all kinds of data and the limitation of bandwidth, optical communication systems based on fiber amplifiers and traditional dense wavelength division multiplexing (DWDM) have gradually failed to meet the rapidly increasing information demand. With the research of polarization multiplexing technology in recent years, two orthogonal polarization states are adopted to transmit information independently, which does not require extra bandwidth, which makes the transmission capacity increase and the spectrum efficiency further improved on the basis of the existing multiplexing technology. Combined with coherent detection technology, it is widely used in modern high-speed optical communication systems. In optical communication systems, with the application of fiber amplifiers, the influence of optical fiber loss on communication quality can be neglected, so at present, people focus on dispersion compensation. At present, more mature dispersion compensation techniques are mainly carried out in the optical domain, and most of them are aimed at direct detection systems, which can no longer meet the requirements of high-speed coherent optical communication. In this paper, the adaptive algorithm used in the digital processing module is studied to solve the problem of the contradiction between the convergence speed and the convergence accuracy of the algorithm, and the algorithm used in the high-order modulation is deeply studied. The main contents of this paper are as follows: (1) in this paper, we do a lot of theoretical analysis on the electric-domain equalization algorithm, deduce the problem of the contradiction between the convergence speed and the convergence time of the algorithm, and introduce two kinds of optimization schemes for this problem. Through the analysis of the improved scheme, it is found that the computational complexity is higher than that of the fixed-step algorithm. In order to solve this problem, this paper proposes an improved piecewise iterative step size algorithm, which includes fixed step size and variable step size. The improved method is applied to the least square method and constant modulus algorithm. The feasibility of the scheme is verified by polarization multiplexing QPSK modulation system. The results show that the improved piecewise step size iterative algorithm is better than the general improved algorithm in convergence speed and accuracy, and the computational complexity of the algorithm is lower than that of the general improved algorithm.) for the 16QAM modulation mode, in order to ensure the high spectral efficiency of the system, The current equalization algorithm is a constant modulus algorithm without reference sequence. However, the number of modules in 16QAM modulation mode is not unique, and the general constant modulus algorithm can not be used directly. An improved algorithm based on virtual and real separate calculation is proposed to solve this problem. This algorithm can reduce the number of modules. However, the module value can not be changed into a unique value, and the computational complexity of the algorithm is high. Therefore, an improved constant modulus algorithm based on decision theory is proposed in this paper, which utilizes the unique characteristics of the received signal modulus value to realize the constant modulus algorithm. Through analysis, the algorithm based on decision theory has simpler structure, better equalization effect and lower computational complexity than virtual and real algorithm.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.1

【参考文献】