Empirical study on human mobility for mobile wireless networks

The knowledge of human mobility is essential to routing design and service planning regarding both civilian and military applications in mobile wireless networks. In this paper, we study the inherent properties of human mobility upon our collected GPS moving traces. We found that power laws characterize the human mobility in both spatial and temporal domains. In particular, because of the diurnal cycle patterns of human daily activities in associated social territories with limited size, there always exists a characteristic distance in the power law distributions of trip displacement and distance between site locations and a characteristic time in the power law distributions of pause and site return time, respectively. Thus, the CCDF of human movement metrics in spatial and temporal domains always has a transition from power-law head to exponential tail delimited by the associated characteristic distance and characteristic time, respectively. Furthermore, we found that either human random moving direction process without pause or the power law distribution of trip displacement lead to a superdiffusive human mobility pattern, while the power law distribution of pause time causes a subdiffusive human movement pattern.