论文部分内容阅读
主要研究移动核心网BSC的TRAU数据分布规律及可疑故障识别方法.由于BSC的最大承载量是未知的,因此本文定义了一类TRAU数据突增故障现象,并在此基础上定义了可疑故障点、异常点和广义异常点.提出并建立了移动核心网故障数据二重联合线性判别模型和三类模式判别准则,用来对移动核心网TRAU数据进行三类模式(“正常”、“异常”、“可疑故障”)识别.实例分析表明,本文所建立的移动核心网故障数据二重联合线性判别模型,当用来对TRAU数据进行三类模式判别时,对原数据样本的回判和新数据样本的识别准确率都达到百分之百,因此能有效识别TRAU数据中的可疑故障点.
This paper mainly studies the rules of TRAU data distribution and suspicious fault identification in mobile core network.Because the maximum capacity of BSC is unknown, this paper defines a phenomenon of sudden burst of TRAU data and defines a suspicious fault point , Abnormal points and generalized anomaly points.Two joint joint discriminant models and three types of discriminant criteria for mobile core network fault data are proposed and established to classify three types of mobile core network TRAU data (“normal”, “ ”Abnormal“, ”suspicious fault") .An example analysis shows that when the dual joint linear discriminant model of mobile core network fault data established in this paper is used to distinguish TRAU data from three types of modes, the original data Sample re-evaluation and new data samples are all 100% accurate, thus effectively identifying suspicious points in the TRAU data.