基于数据辅助线性相位估计的超高速光波分相干传输关键技术研究

Coherent optical signal detection in ultra high capacity transmission systems has the potential on effective spectral efficiency of novel optical modulation formats, high receiver sensitivity and tolerance to accumulated residual dispersion, etc. Since it is difficult to implement high speed optical phase-locked loop, extraction of optical carrier phase information for synchronous coherent detection can only be replaced by optical carrier phase estimation. Beneficial from recent rapid development of high speed electrical digital signal processing techniques, such a novel scheme is a quasi-ideal optical coherent detection method. However, real-time coherent detection is still difficult to be implemented because of complicated nonlinear mathematical operation in the process of such carrier phase estimation. We first proposed a novel decision-aided maximum likelihood (DA-ML) optical carrier phase estimation scheme in 2008, which is based on a simple linear arithmetic algorithm to ensure the potential of non-offline processing on coherent optical signal detection. Previous pioneering works are mainly focus on different versions of single carrier receiver design with theoretical performance evaluation, and the most important results have been reported in recent SPIE newsroom and an OFC 2012 invited paper. In this project, we prepare to further investigate performance optimization of such a DA-ML scheme in wavelength division multiplexing optical coherent transmission system from the aspect of fundamental theory and practical application with experimental verifications so as to exploring the future online coherent detection. Our research mainly includes performance evaluation and advanced detection schemes for systems under different dispersion maps and amplification schemes so as to counteract various nonlinear deteriorations from the crosstalk of neighboring channels. We also target to the reasonable characteristic requirement for a practical high speed digital sampling system with achievable filtering bandwidth and sampling precision in a DA-ML coherent receiver.

超高传输容量光相干信号检测用以挖掘新型光调制格式的高频谱利用率、高接收灵敏度和色散容忍度等方面的潜力。因难以实现高速光锁相环，同步检测必需的光载波相位信息提取只能由相位估计技术替代，这是近年来利用高速电信号数字处理技术实现的准理想化光相干检测方案，但相位估计过程中复杂非线性运算使得实时信号检测困难。2008年我们首次提出了辅助判决式最大似然光载波相位估计技术，完全基于线性化算法，确保高速光信号实时相干检测。前期工作主要集中在单信道接收机方案的理论，相关研究成果受邀在SPIE newsroom专题报道及OFC 2012特邀报告。本项目拟进一步研究线性载波相位估计方案在光波分复用系统中的性能与优化，解决理论、实验和应用方面的关键问题，为在线相干检测奠定基础。研究内容主要包括各种色散分布、放大方案和相邻信道非线性特性前提下的检测方案与特性，并探讨实用高速数字采样系统滤波带宽和采样精度要求等。

辅助判决式最大似然(DA-ML)光载波相位估计是一种可用于相干光通信系统中完全基于线性化的相位估计方案，用以确保高速光信号实时相干检测。项目研究着重解决该方案内在的问题与不足，开展对DA-ML相干光接收机的改进型策略进行研究。项目研究最主要的贡献在于，提出了全新的灵活 DA-ML接收机方案, 该方案基于相对较长的传输序列，考虑相邻码元的相位缓变问题，可以很好地对幅度噪声平均，解决了传统方案在低速单载波及OFDM多载波系统中遇到的应用难题，同时根本消除了必需预设或自适应获得最佳块长度问题，系统在开始时仅需选择较长符号块工作，即可以达到准最佳工作状态。研究工作同时深入考察系统中存在的其他关键问题，例如，高速模数转换器件的采样精度或最低分辩率的要求，信号处理器与接收电路可能存在的最佳带宽范围问题，接收机冷启动策略及未来突发模式相干接收，混合光电色散补偿系统中非线性传输及增强噪声，等等。项目研究内容、进程与目标与计划书基本一致，同时与产业界的需求完全一致，项目成果远超预期的要求。在此基础上，项目研究拓宽对高速信号处理技术在光通信中的应用基本问题的研究，特别是探索在可见光通信系统中的应用，首次提出了基于预编码算法的多用户多入多出系统模型，通过对发送信号的预处理，在发送端将多用户干扰消除，从而有效降低用户终端的复杂度和功耗。项目研究工作同时提出了OFDM帧内信号时域调整方案，以及多脉冲位置调制式调光用以传输额外信息的方案等等。项目研究这一在无线光通信中拓宽性探索，对于追踪研究光通信业界最新需求与进展有着相当大的现实意义，符合自然科学基金资助的进一步科学探索意义。