摘要
传统的运维模式已无法满足4G以及5G网络运维的需要,运用人工智能和大数据等新技术,实现主动运维、快速运维和精准运维已成为网络运维发展的新趋势。基站异常检测是新型运维模式的一项重要内容,因此提出一种基于深度自编码器的基站异常度检测方法。首先对故障告警、相关性能KPI以及OMC运维指标分别建立稀疏降噪自编码器模型,然后综合三个模型的结果,对基站进行全面的异常度评测。该方法具有准确性高、评测粒度细、容易实施等特点,经过实际的试点应用,验证了该方法的有效性,为后续网络运维部门进行基站精准巡检以及进一步实现智能运维提供了可靠的数据支撑。
The traditional operation and maintenance(O&M)mode has not been able to meet the requirements of 4G and 5G network O&M.It has become a new development trend of network O&M with new technologies,such as artificial intelligence,big data,to realize active O&M,fast O&M and precise O&M.Base station anomaly detection is an important part of the new O&M mode,and this paper propose an anomaly detection method for base stations based on deep auto-encoder.Firstly,three sparse de-noising auto-encoder(SDAE)models are established with fault alarms,key performance indicators(KPIs)and OMC O&M indicators,respectively.Then a comprehensive anomaly evaluation of the base station is given on the combination of the results of the three models.The proposed method has the characteristics of high accuracy,fine evaluation granularity,and easy implementation.Through pilot applications,the effectiveness of the method is verified.Therefore,it provides reliable data support to precise inspection of base station and further realization of intelligent O&M.
作者
马敏
贾子寒
王磊
MA Min;JI AN Zihan;WANG Lei(Shaanxi Branch,China Mobile Group Design Institute Co.,Ltd.,Xi'an 710065,China;China Mobile Group Design Institute Co.,Ltd.,Beijing 100080,China)
出处
《移动通信》
2021年第5期124-129,134,共7页
Mobile Communications
关键词
基站异常检测
深度学习
自编码器
稀疏降噪
基站精准巡检
智能运维
base station anomaly detection
deep learning
auto-encoder
sparse de-noising
base station precise inspection
intelligent operation and maintenance
