关键词: 数据挖掘, 聚类分析, 划分, 基于密度的带有噪声的空间聚类（DBSCAN）, 载频优化, 最佳话务量

Abstract: In order to take full advantage of the wireless network resources and improve the quality of municipal wireless networks, this paper makes full use of DBSCAN（Density Based Spatial Clustering of Applications with Noise） algorithm and proposes a traffic abnormal cells detection method. Some statistical analysis of the existing network traffic data has been done to identify the relationship between carrier configuration and the best traffic. This article optimizes the cell’s carrier configuration, which has unusual traffics and high congestion rates. It also has guiding meaning for city cell wireless network optimization.

Key words: data mining, clustering, partition, Density Based Spatial Clustering of Applications with Noise（DBSCAN）, carrier optimization, best traffic