一种基于K-means算法的无线业务分布预测方法

he traffic distribution forecasting is part of the wireless network planning. According to the traffic distribution, the mobile network can get the accurate network capacity and coverage; the traffic distribution also plays an important role on the early stages of planning and the latter network optimization. This paper first introduces the common algorithms of Traffic Distribution Forecasting in mobile communication network and analyzes their advantages and disadvantages, then gives a new effective algorithm based on cluster analysis of K-means. In this new algorithm the cluster analysis characters of traffic distribution are extracted, which are based on digital map and traffic data and calculates the traffic distribution in a class have the similar characters. Apparently, to compare with other common algorithms, this algorithm enhances the credibility of traffic distribution forecasting.