About the relationship between people and discoverable Bluetooth devices in urban environments

Abstract

People are frequently carrying their Bluetooth enabled mobile phones in their pockets while moving through public spaces. With only a small fraction of these being set to discoverable mode, it is possible to estimate the number of persons in proximity by conducting Bluetooth device inquiries. In this paper, we present a method for the measurement of the percentage of people with discoverable devices. We did experiments using this method in Bremen, Germany and San Francisco, US, suggesting that about 2%, respectively 6%, are detectable. Further, we show that this measurement can well be implemented on common mobile phones, and that specialized scanners are not necessary.

Bluetooth based collaborative crowd density estimation with mobile phones

“…The work most similar to ours is by Nicolai et al [2] where the discovery time of Bluetooth devices as well as the relation between the number of people and the number of discoverable Bluetooth devices was investigated. As opposed to our approach the work relied on static Bluetooth sensing locations and only the absolute number of discovered Bluetooth devices was used.…”

“…The instantaneous number of discoverable Bluetooth devices in the covered area is mapped to a number of people by assuming a fixed proportion between the number of discoverable Bluetooth devices and the number of people. [2] Collaborative and in combination of stationary and dynamic measurements (crowd density) This novel approach is considered in this paper. With multiple people equipped with ubiquitous smartphones we present new features to evaluate the crowd-density instantaneously.…”

Opportunistic Sampling for Joint Population Size and Density Estimation

“…Several of these counts can be performed at different time periods and the resulting ratios averaged [15]. The same idea is used in the works of [40], [41] to propose techniques to estimate the ratio of discovered Bluetooth devices. Let N Tot be the total number of attendees and let r be the ratio of attendees carrying visible Bluetooth devices, i.e., r ¼ N=N Tot .…”

CrowdWatch

“…Mobile phones have been used to estimate density by scanning the environment for discoverable Bluetooth devices [14,15]. Ideally, all nodes monitor their neighborhood all of the time.…”