Self-adaptive Trajectory Prediction Method Based on Density Clustering

To solve the problem of user trajectory prediction in the mobile communication environment, we proposed a self-adaptive trajectory prediction method (ATPDC) based on density clustering. And it consists of two stages which are trajectory modeling stage and trajectory updating stage respectively. In the first stage, it constructs the user trajectory prediction model by clustering historical trajectory. And in the second stage, it enhances the model built on the former stage. We test it on the MR records in the mobile communication environments. Experimental results show that ATPDC algorithm can achieve the incremental updating with satisfactory prediction accuracy and prediction efficiency with the growth of user trajectory data. Furthermore, it is also suggested that the mobile MR road test reports contain potential user behavior patterns and could be used to analyze and mine users’ behaviors.