With the development of wireless communication technology, the utilization of the ambient information that the network users observe has attracted much attention. In this paper, we focus especially on the utilization of observed environmental signals for authentication systems. Ambient information taken as unique data at a particular time and place can be utilized to construct stronger authentication systems. However, since the environmental condition of the network is different for each location, which has a huge effect on the observed ambient information in the area, required data are also different for each place. Thus, in this paper, we propose an efficient data collection method which dynamically changes the way data is collected according to the requirements and the network condition. More specifically, our proposal aims to collect uniform amount of fresh data from areas with varying population density. Additionally, an algorithm to improve the efficiency of our proposal regarding the accommodation of the environmental condition of the network is introduced. Moreover, numerical results verify the effectiveness of our proposal.