Detect User Transition Areas on Community Structure with Noise in 5G Mobile Environment

Mintao Liu, Tunzi Tan, Kangyu Shang, Suixiang Gao, Wenguo Yang, Yang Yang


Detecting communities problems have been studied under various background and networks in last two decades. This paper poses a new problem, named community detection with noise, to find communities where nodes connect densely to each other under the background of user movement in 5G. Because the prospective small cells in 5G would make data lengthy [1] [2], this paper aims to figure out meaningful areas without seldom visited cells named as “noise”. Two kinds of methods are proposed. One is based on the normal community detection problem. Noise nodes are picked and deleted by proper quantities. The other is a heuristic algorithm that starts with the most important node, adding closely connected nodes into the community. Compared with classic GN algorithm, these two methods outperform in quality metrics that focus on internal density. And a practical example that gives user movement areas is showed.


Full Text:



  • There are currently no refbacks.