人工智能技术的高速光纤通信链路智能选择研究  

Research on intelligent selection of high-speed optical fiber communication links based on artificial intelligence technology

作　　者：林逢春 谭文武 LIN Fengchun;TAN Wenwu(Jiangxi college of engineering,Xinyu Jiangxi 338000,China)

机构地区：[1]江西工程学院,江西新余338000

出　　处：《激光杂志》2022年第12期159-163,共5页Laser Journal

基　　金：江西省教育厅科学技术研究项目(重点项目)(No.191185)。

摘　　要：随着光纤网络规模的不断扩大,网络中的数据不断增加。当前光纤通信链路智能选择方法时常由于信号传输预测能力的不足,造成链路带宽损失过大的问题,为此提出子人工智能技术的高速光纤通信链路智能选择方法。根据链路非线性效应原理,完成光纤链路噪声处理。选用扩展卡尔曼滤波算法,对通信网络信号走向进行预测。根据强化学习算法,对多个链路选择后通信质量展开验证,获取最优通信链路。构建测试实验环节,对其使用效果进行验证。经实验结果可知,该方法可弥补当前方法在使用中的不足,提高光纤网络的通信质量。With the continuous expansion of the scale of optical fiber networks,the data in the network is increasing.At present,the intelligent selection method of optical fiber communication link often causes the problem of excessive link bandwidth loss due to the lack of signal transmission prediction ability.Therefore,an intelligent selection method of high-speed optical fiber communication link based on artificial intelligence technology is proposed.According to the principle of link nonlinear effect,the noise processing of the optical fiber link is completed.The extended Kalman filter algorithm is selected to predict the signal trend of the communication network.According to the reinforcement learning algorithm,the communication quality of multiple links is verified to obtain the optimal communication link.Construct the test experiment link to verify its application effect.The experimental results show that this method can make up for the shortcomings of the current methods and improve the communication quality of optical fiber networks.

关 键 词：人工智能技术 光纤通信网络 卡尔曼滤波算法 链路智能选择 带宽损失率 

分 类 号：TN929.12[电子电信—通信与信息系统]