论文部分内容阅读
在高速光纤偏分复用(PDM)16进制正交幅度调制(QAM)相干光通信系统中,偏分解复用算法是分离偏振信息和进行信号补偿的关键技术。针对传统盲解复用的恒模算法(CMA)易于陷入奇异性和独立成分分析(ICA)算法收敛性能一般的问题,本文提出了一种改进型ICA(MICA)算法。算法基于给出的一种全新收敛测量方法,通过计算补偿后信号点与理想星座点的距离,自适应的改变学习速率,达到了更好的误码率(BER)性能。在20Gbaud/s的PDM-16-QAW系统上的仿真结果表明,与传统CMA法相比,本文提出的算法完全消除了奇异性问题,与是否存在偏振相关损耗(PDL)无关。此外,算法还能在前向纠错编码门限上提供高达1dB的光信噪比(OS-NR)提升,并在收敛速率和精确度上都得到了提升。
In PDM QAM coherent optical communication systems, partial decomposition and demultiplexing algorithm is the key technology to separate the polarization information and perform signal compensation. In order to solve the problem that traditional CMA is easy to get into the problem of convergence of ICA algorithm, this paper presents an improved ICA (MICA) algorithm. The proposed algorithm, based on a new convergence measurement method, achieves better BER performance by calculating the distance between the compensated signal point and the ideal constellation point and adaptively changing the learning rate. The simulation results on the 20Gbaud / s PDM-16-QAW system show that the proposed algorithm completely eliminates the singularity problem compared with the traditional CMA method, regardless of the presence or absence of polarization-dependent loss (PDL). In addition, the algorithm provides up to 1dB of optical signal-to-noise ratio (OS-NR) improvement at the forward error correction coding threshold, with an improvement in convergence rate and accuracy.