Author Login Editor-in-Chief Peer Review Editor Work Office Work

Computer Engineering ›› 2021, Vol. 47 ›› Issue (3): 83-93. doi: 10.19678/j.issn.1000-3428.0057318

• Artificial Intelligence and Pattern Recognition • Previous Articles     Next Articles

Recognition Algorithm for Space-Time Density Track Points of Celluar Signaling

CHEN Lüe, XIONG Chen, CAI Ming   

  1. Guangdong Province Key Laboratory of Intelligent Transportation Systems, School of Intelligent Systems Engineering, Sun Yat-Sen University, Guangzhou 510006, China
  • Received:2020-02-05 Revised:2020-03-11 Published:2020-03-16

手机信令的时空密度轨迹点识别算法

陈略, 熊宸, 蔡铭   

  1. 中山大学 智能工程学院 广东省智能交通系统重点实验室, 广州 510006
  • 作者简介:陈略(1994-),女,硕士研究生,主研方向为手机信令数据挖掘与分析;熊宸,副研究员、博士;蔡铭(通信作者),教授、博士。
  • 基金资助:
    国家重点研发计划(2018YFB1601001);高校基本科研资助项目(19lgpy290)。

Abstract: Celluar signaling is spatiotemporal sequential,and has some features such as large amount of data,uneven sampling frequency,low positioning accuracy and base station oscillation,which lead to the uneven distribution of data density,large space-time overhead and poor clustering effect of traditional Celluar signaling clustering methods.To address the problem,this paper proposes an algorithm to recognize the track points of space-time density of celluar signaling.The celluar signaling data is gridded to unify the evaluation scale.According to the characteristics of oscillatory noise,the grid clusters are spatiotemporal connected to reduce the spatial uncertainty and the amount of computation.Combined with the tortuosity of the network trajectory and the movement and residence time,the spatiotemporal movement ability of the trajectory points in the grid cluster is redefined.The spatiotemporal density of the grid cluster is calculated to judge the user's residence area,and the mobile residence tags Trajectory data are collected to verify the effectiveness and recognition efficiency of the algorithm.Experimental results show that the recognition accuracy of this algorithm is higher than that of the improved DBSCAN algorithm,which is suitable for identifying the residence area of celluar signaling data, and the recognition effect of complex trajectory residence area is better.

Key words: celluar signaling, space-time connection, space-time mobility, space-time density, residence area

摘要: 手机信令具有时空序列性以及数据量大、采样频率不均、定位精度低与基站振荡等特点,导致传统手机信令聚类方法数据密度分布不均、时空开销大且聚类效果差。提出一种用于手机信令的时空密度轨迹点识别算法。将手机信令数据网格化以统一评估尺度,根据振荡噪声特征对网格簇进行时空联结减少空间不确定性和计算量,结合网络轨迹的曲折性以及移动与停留时间重新定义网格簇内轨迹点时空移动能力,计算网格簇的时空密度以判断用户停留区域,并采集具有移动停留标签的轨迹数据以验证算法有效性和识别效率。实验结果表明,该算法识别精度较改进DBSCAN算法更高,适用于识别手机信令数据停留区域,对复杂轨迹停留区域的识别效果更好。

关键词: 手机信令, 时空联结, 时空移动能力, 时空密度, 停留区域

CLC Number: