基于卷积神经网络的无线视频监控干扰识别算法研究  

A Study on Interference Identification Algorithm for Wireless Video Surveillance Based on Convolutional Neural Network

作　　者：葛澍 王琳 高峰 李行政 GE Shu;WANG Lin;GAO Feng;LI Xingzheng(China Mobile Communications Group Co.,Ltd.,Beijing 100032,China;China Mobile Group Design Institute Co.,Ltd.,Beijing 100080,China)

机构地区：[1]中国移动通信集团有限公司,北京100032 [2]中国移动通信集团设计院有限公司,北京100080

出　　处：《移动通信》2023年第7期104-110,共7页Mobile Communications

摘　　要：在5G网络干扰优化过程中发现,目前2.6 GHz频段面临无线视频监控干扰问题,严重影响5G网络质量。从问题产生的根本原因出发,结合一线干扰排查经验提出了一种融合干扰信号特征的卷积神经网络无线视频监控干扰识别算法,并使用现网数据进行算法验证,证明可快速、精准识别受视频监控干扰的5G小区,通过网管数据分析为干扰定位提供重要的先验信息,提升此类小区干扰定位效率。During the process of optimizing interference in 5G networks,the challenge of wireless video surveillance interference in the 2.6 GHz frequency band has been recognized as a critical issue,with a severe impact on the quality of the 5G network.This research aims to investigate the underlying causes of this problem and proposes an advanced algorithm for wireless video surveillance interference recognition that integrates the distinctive characteristics of interference signals using convolutional neural networks.Rigorous validation using authentic network data substantiates the algorithm's exceptional capability to swiftly and accurately identify the specific 5G cells that are adversely affected by video surveillance interference.Additionally,by analyzing network management data,the algorithm provides important prior information for interference localization,thereby improving the efficiency of interference localization in such cells.

关 键 词：无线视频监控干扰 干扰识别 5G 

分 类 号：TN915[电子电信—通信与信息系统]