相干检测中的DSP算法和仿真研究
发布时间：2019-05-14 19:28
【摘要】：相干检测是一种全息检测技术,它可以允许信息被编码在光纤的所有自由度上,包括幅度,相位以及偏振。因此,它可以提高信号功率以及系统的频谱效率。近年来,由于高清电视、视频会议等新业务的驱动,网络容量迅速增长,之前的带宽已经不够使用,具有高频谱效率的相干检测技术开始受到人们的广泛关注。 同时,高速率模数转换器(ADC)的实现以及大规模集成电路(VLSI)的发展使得对光纤传输损伤在GHz波特速率上进行数字补偿成为可能。因此,与数字信号处理(DSP)技术的结合将使相干检测技术更加具有吸引力。对于相干检测后的信号,DSP算法可以对光纤信道的线性效应做到无损补偿,例如色散效应和偏振模色散效应。此外,DSP算法也可以实现发射端激光器与本地激光器的载波同步。基于以上原因,对于相干检测技术中DSP算法的研究非常火热。 本文首先介绍了基于DSP的相干光系统的基础知识,然后介绍了目前应用在相干光通信中的DSP算法,包括色散补偿、解偏振串扰、非线性补偿、频率估计以及相位估计算法五个方面。这些算法普遍存在着复杂度大的缺点,为了降低硬件功耗,复杂度必须要降低。作者在相位估计和频率估计这两个方面做出了改进。对于频率估计中的共轭M次方算法,它需要采用复数乘法器来去除调制相位信息,这会带来硬件功耗的增大。基于此,作者提出了一种低复杂度的M次方算法,这种算法采用查表器件来去除调制相位信息。与传统的共轭M次方算法的性能一样的同时,这个算法不需要采用复数乘法器,因而使复杂度大大降低。此外,作者还在相位估计算法方面做出了创新。盲相位搜索算法是目前相位估计中精度最高的算法之一,它首先假设出所有可能的相位噪声值,然后再借助欧式距离最短原则选出最接近于真实相位噪声的测试相位。为了保证盲相位搜索算法的高精度,在盲相位搜索算法中一般采用逐符号估计,同时测试相位个数也要求很多,这会带来很大的复杂度。在本文中,基于相位噪声慢变的特性,作者提出了粗估计和细估计相结合的两步法以及滑动块的思想,降低了测试相位的个数以及每个符号重复计算的次数。通过采用两步法以及滑动块的思想,原始的盲相位搜索算法的复杂度大大降低。仿真结果表明,在激光器线宽容忍度降低11%的同时,算法复杂度至少可降低一半。
[Abstract]:Coherent detection is a holographic detection technique, which allows information to be encoded on all degrees of freedom of optical fiber, including amplitude, phase and polarization. Therefore, it can improve the signal power and the spectral efficiency of the system. In recent years, due to the drive of high-definition television, video conferencing and other new services, the network capacity has increased rapidly, and the bandwidth has not been used enough. the coherence detection technology with high spectral efficiency has been paid more and more attention. At the same time, the implementation of high rate analog-to-digital converter (ADC) and the development of large-scale integrated circuit (VLSI) make it possible to digitally compensate the optical fiber transmission damage at GHz port rate. Therefore, the combination with digital signal processing (DSP) technology will make coherent detection technology more attractive. For coherent signals, DSP algorithm can compensate the linear effect of fiber channel nondestructive, such as dispersion effect and polarization mode dispersion effect. In addition, DSP algorithm can also realize carrier synchronization between emitter laser and local laser. For the above reasons, the research on DSP algorithm in coherent detection technology is very hot. In this paper, the basic knowledge of coherent optical system based on DSP is introduced, and then the DSP algorithm used in coherent optical communication is introduced, including dispersion compensation, depolarization crosstalk, nonlinear compensation. There are five aspects of frequency estimation and phase estimation algorithm. These algorithms generally have the disadvantage of large complexity. In order to reduce the hardware power consumption, the complexity must be reduced. The author has made improvements in phase estimation and frequency estimation. For the conjugated M-th power algorithm in frequency estimation, it needs to use the complex multiplier to remove the modulation phase information, which will lead to the increase of hardware power consumption. Based on this, a low complexity M-power algorithm is proposed, which uses look-up table devices to remove modulation phase information. The performance of this algorithm is the same as that of the traditional conjugated M-th power algorithm, at the same time, this algorithm does not need to use complex multiplier, so the complexity is greatly reduced. In addition, the author also makes an innovation in the phase estimation algorithm. Blind phase search algorithm is one of the most accurate algorithms in phase estimation at present. It first assumes all possible phase noise values, and then selects the test phase which is closest to the real phase noise with the help of the principle of shortest Euclidean distance. In order to ensure the high accuracy of blind phase search algorithm, symbol by symbol estimation is generally used in blind phase search algorithm, and the number of test phases is also required, which will bring great complexity. In this paper, based on the slow change of phase noise, the author proposes a two-step method combining rough estimation with fine estimation and the idea of sliding block, which reduces the number of test phases and the number of repeated calculations per symbol. By using the two-step method and the idea of sliding block, the complexity of the original blind phase search algorithm is greatly reduced. The simulation results show that the complexity of the algorithm can be reduced by at least half while the tolerance of laser line is reduced by 11%.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TN26;TP368.1
本文编号：2476987
[Abstract]:Coherent detection is a holographic detection technique, which allows information to be encoded on all degrees of freedom of optical fiber, including amplitude, phase and polarization. Therefore, it can improve the signal power and the spectral efficiency of the system. In recent years, due to the drive of high-definition television, video conferencing and other new services, the network capacity has increased rapidly, and the bandwidth has not been used enough. the coherence detection technology with high spectral efficiency has been paid more and more attention. At the same time, the implementation of high rate analog-to-digital converter (ADC) and the development of large-scale integrated circuit (VLSI) make it possible to digitally compensate the optical fiber transmission damage at GHz port rate. Therefore, the combination with digital signal processing (DSP) technology will make coherent detection technology more attractive. For coherent signals, DSP algorithm can compensate the linear effect of fiber channel nondestructive, such as dispersion effect and polarization mode dispersion effect. In addition, DSP algorithm can also realize carrier synchronization between emitter laser and local laser. For the above reasons, the research on DSP algorithm in coherent detection technology is very hot. In this paper, the basic knowledge of coherent optical system based on DSP is introduced, and then the DSP algorithm used in coherent optical communication is introduced, including dispersion compensation, depolarization crosstalk, nonlinear compensation. There are five aspects of frequency estimation and phase estimation algorithm. These algorithms generally have the disadvantage of large complexity. In order to reduce the hardware power consumption, the complexity must be reduced. The author has made improvements in phase estimation and frequency estimation. For the conjugated M-th power algorithm in frequency estimation, it needs to use the complex multiplier to remove the modulation phase information, which will lead to the increase of hardware power consumption. Based on this, a low complexity M-power algorithm is proposed, which uses look-up table devices to remove modulation phase information. The performance of this algorithm is the same as that of the traditional conjugated M-th power algorithm, at the same time, this algorithm does not need to use complex multiplier, so the complexity is greatly reduced. In addition, the author also makes an innovation in the phase estimation algorithm. Blind phase search algorithm is one of the most accurate algorithms in phase estimation at present. It first assumes all possible phase noise values, and then selects the test phase which is closest to the real phase noise with the help of the principle of shortest Euclidean distance. In order to ensure the high accuracy of blind phase search algorithm, symbol by symbol estimation is generally used in blind phase search algorithm, and the number of test phases is also required, which will bring great complexity. In this paper, based on the slow change of phase noise, the author proposes a two-step method combining rough estimation with fine estimation and the idea of sliding block, which reduces the number of test phases and the number of repeated calculations per symbol. By using the two-step method and the idea of sliding block, the complexity of the original blind phase search algorithm is greatly reduced. The simulation results show that the complexity of the algorithm can be reduced by at least half while the tolerance of laser line is reduced by 11%.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TN26;TP368.1
【参考文献】
相关期刊论文 前1条
1 李新;顾畹仪;兰名荥;;相干光检测系统中的DSP技术研究[J];光通信技术;2011年08期
相关博士学位论文 前1条
1 丁攀峰;单模光纤中偏振效应的理论研究及分析[D];华中科技大学;2007年
相关硕士学位论文 前2条
1 曹寅文;相干检测中的载波同步DSP算法和仿真[D];北京邮电大学;2011年
2 刘向春;偏振模色散模拟和色散补偿的研究[D];北京交通大学;2008年
，本文编号：2476987
本文链接：https://www.wllwen.com/kejilunwen/jisuanjikexuelunwen/2476987.html
