人工智能技术的复杂光网络信号调制方式分类研究  

Research on classification of signal modulation modes in complex optical networks based on artificial intelligence technology

作　　者：毛丽平[1] 张建辉[1] MAO Liping;ZHANG Jianhui(Nanchang Institute of Technology,Nanchang 3305111 China)

机构地区：[1]南昌理工学院,南昌330511

出　　处：《激光杂志》2022年第8期107-111,共5页Laser Journal

基　　金：教育部科技发展中心高校产学研创新基金(No.2018A05031)。

摘　　要：为了提高复杂光网络信号检测识别能力,需要进行信号调制方式分类,提出基于人工智能技术的复杂光网络信号调制方式分类与识别方法。构建复杂光网络信号的时域模型,通过频域参数识别和信道扩频序列定位方法实现对复杂光网络信号的特征检测,提取信道扩频序列调制参数,通过人工智能算法构建复杂光网络信号的人工智能分类结构模型,对提取的复杂光网络信号调制类别特征分量进行自适应融合和分类识别。仿真结果表明,本方法调制分类的平均精度为93.46%,分类识别误差平均为2.79%,耗费的时间平均为11.31 s,具有较好的复杂光网络信号调制方式分类性能。In order to improve the detection and recognition ability of complex optical network,it is necessary to classify the modulation mode of the signal.A method of signal modulation classification and recognition based on artificial intelligence technology is proposed.The time domain model of complex optical network signal is constructed.The feature detection of complex optical network signals is realized by frequency domain parameter identification and channel spread spectrum sequence positioning method.The modulation parameters of the channel spread spectrum sequence are extracted,and the artificial intelligence classification structure model of complex optical network signals is constructed by artificial intelligence algorithm,The extracted modulation feature components of complex optical network are fused and classified.Simulation results show that the average accuracy of modulation classification is 93.46%,the average error of classification recognition is 2.79%,and the average time spent is 11.31 s,which shows a good classification performance of complex optical network signal modulation methods.

关 键 词：人工智能技术 复杂光网络信号 调制方式 分类 识别 

分 类 号：TN919[电子电信—通信与信息系统]