基于决策树的TD-LTE基站规划方法研究

摘要

分析了以TD-LTE基站的最近站间距、平均站间距、站高、第1层邻区数作为分类的特征数据，以每个基站实际测试的下载速率为优化目标，构建了决策树算法，将基站类型分为5类。通过实际测试数据验证表明，该方法的精度较高，从而为TD-LTE建网初期新站规划提供了参考。%The feature data of TD-LTE BS based on the nearest base station (BS) distance, the average BS distance, the BS height and the number of the adjacent cell of the first layer are analyzed in this paper. To optimize the actual download rate of BS, a decision-tree algorithm is built to classify BSs into five types. The actual test data validate that the proposed algorithm has high precision to provide a reference to new BS construction in the early stage of TD-LTE network.