基于分段序列联合学习网络的光收发机IQ损伤监测  

Optical transceiver IQ impairments monitoring based on segmented sequence joint learning network

作　　者：周以琳 雷印杰[1] Zhou Yilin;Lei Yinjie(College of Electronics and Information Engineering,Sichuan University,Chengdu 610065,China)

机构地区：[1]四川大学电子信息学院,成都610065

出　　处：《现代计算机》2023年第7期30-36,共7页Modern Computer

基　　金：国家自然科学基金(62276176):基于点云-语言交互和任务增量学习的三维视觉推理研究。

摘　　要：光收发机同相/正交损伤监测是光通信系统实现可靠传输不可或缺的技术。提出了分段序列联合学习网络来监测相干光收发机中的IQ损伤。SSLN通过处理分析经DSP处理后的接收信号,找到数据中隐含的发射机(Tx)IQ损伤特征和接收机(Rx)的IQ损伤特征,建立映射关系。对100 km标准单模光纤上的100-Gbaud QPSK/64QAM信号传输进行了数值模拟,以评估SSLN的性能。数值模拟结果表明,当多个损伤同时存在时,所提出的SSLN能够准确有效地监测Tx和Rx的IQ损伤。Optical transceiver In-phase/Quadrature(IQ)impairments monitoring is an indispensable technology for reliable transmission in optical communication systems.This paper proposes a Segmented Sequence Joint Learning Network(SSLN)to monitor IQ impairments in coherent optical transceiver.SSLN processes and analyzes the received signal after DSP processing,finds the IQ impairment characteristics of the transmitter(Tx)and the IQ impairment characteristics of the receiver(Rx)implied in the data,and establishes a mapping relationship.A numerical simulation of 100-Gbaud QPSK/64QAM signaling on a 100 km standard single-mode fiber was performed to evaluate the proposed method in this paper.Numerical results demonstrate that the proposed SSLN can accurately and effectively monitor IQ impairments in both Tx and Rx when multiple impairments co-exist.

关 键 词：光收发机 IQ损伤监测 数字信号处理 分段序列联合学习网络 深度学习 

分 类 号：TN929.1[电子电信—通信与信息系统]