TD-LTE移动通信设备故障诊断算法研究

Research on Fault Diagnosis Algorithm of TD-LTE Mobile Communication Equipment

针对TD-LTE移动通信设备中故障样本,提出一种基于相似度与支持向量机(Support Vector Machine,SVM)相融合的移动通信设备故障诊断算法(Similarity Fusion Based Support Vector Machine in Communication Equipment System,SFBSVM),对移动通信设备系统建模和未标记的故障样本进行初始聚类,构造出最终分类器。算法能减少标记样本的数目,降低初选样本对分类器的影响,抑制孤立样本点对分类结果的影响。实验结果表明,SFBSVM计算简单,精度更高,准确率较高且稳定可靠。

Aiming at fault sample in TD-LTE mobile communication equipment, this paper puts forward a Fault Diagnosis Algorithm of Similarity Fusion Based Support Vector Machine in Communication Equipment System（ SFBSVM） ,which performs model-ing for mobile communication equipment system and initial clustering for unlabeled fault samples to get the final classifier.The SFBSVM can reduce the number of labeled samples,the effect of classifier on primary samples,and inhibit the effect of isolated sample point on the classification results.The theoretical analysis and experimental results show that the SFBSVM have such advantages as simple calcu-lation,higher accuracy,higher accuracy rate,higher stability and higher reliability.