Hierarchical of network structure is very important for studying LTE network performance. However

in the actual landing process of sites

due to various reasons

some of sites hierarchical positioning is relatively vague. In this paper

a hierarchical algorithm of network structure through big data analysis and unsupervised clustering methods is proposed and the results obtained by this algorithm are compared with the results of artificial classification of sampling cells. The experimental result shows that the hitting rate of the algorithm is high

which has great reference value for the next step hierarchical of network structure research.