PDM-16QAM系统中几种相位估计算法的性能与复杂度比较以及改进

在偏分复用的高速相干光通信系统中,16进制正交幅度调制(PDM-16QAM)是最有前景的调制格式,且载波相位估计是补偿发射端与接收端激光器对信号造成的相位失真的重要技术。该领域最经常使用的算法为QPSK分区算法与盲相位搜索算法,但他们都有自己的缺点,如估计效果不够好,适用范围不够广,或者计算复杂度过高。本论文实验搭建了224Gb/s PDM-16QAM波分复用传输640km相干检测系统平台,首先利用几种基于前述两种算法的相位估计方法对接收到的信号进行了数据处理,分别给出了误码率曲线图、误差向量幅度曲线图、计算复杂度分析表。然后基于性能最优与计算复杂度的平衡,提出了针对16QAM信号星座图中间圈的点的省略两级处理算法。将该改进的算法用于实验系统相位估计单元进行数据处理,结果表明,改进的算法P1,3/ML其误码率性能和误差向量幅度性能基本保持不变,但是计算复杂度降低为原来的1/2。

Comparison and improvement of performance and complexity of several phase estima tion algorithms in PDM-16QAM system

In high-speed optical polarization division multiplexed (PDM) coherent transmis sion system,16quadrature amplitude modulation (16QAM) is the most promisin g modulation format and the carrier phase estimation is an important technique to compensate the signal phase distor tion caused by the linewidth of lasers in the transmitter and the receiver.The most commonly used methods in the area are QPSK partitioning algorithm and blind phase search (BPS) algorithm,but they suff er from either high hardware consumption,bad performance or narrow range of application.In this pa per,a 640km transmission experimental platform for 224Gbit/s PDM-16QAM WDM co herent detection is set up.First,the received signal is processed by several phase estimation algorithms based on QPS K partitioning algorithm or BPS algorithm,obtaining the bit error rate (BER) curves,the error vector magnitude (EVM) curves and the computational complexity (CC) analysis tables.Then,a modified al gorithm P_1,3/ML omitting the two-stage processing of points in the intermediate ring of 16QAM c onstellation is proposed, which is based on the trade-off between performance and computational comp lexity.After using the modified algorithm in the phase estimation module of the experimental system for data processing,the results show that the computational complexity is reduced to a factor of 2while the BER and EVM performance remain nearly unchanged.