Mobility models play a crucial role in mobility wireless networks with respect to evaluating the network protocols and performances. However, the majority of existing mobility models either does not exhibit realistic movement characteristics or modeling methods are too complex. In this paper, a new mobility model based on Location Attraction (LAMM) is proposed, which utilizes human's clustering features in real life. Through analyzing real GPS trace of mobile users, we observe the number and location attraction of hot regions. We find that the number of hot regions is extremely stationary in an observation area, and the pause position density within hot regions show a trend with exponential decline. Based on such movement characteristics, we developed a mobility model and the validation shows that LAMM better depict the mobility patterns of human.
展开▼
