Group monitoring in mobile wireless sensor networks

The need to monitor groups of mobile entities arises in many application contexts. Examples include the study of the social behavior of humans and wildlife, the shepherding of livestock, the care giving to people that are not self-sufficient. Human- or animal-borne wireless devices can be used to detect the joining or leaving of group members, even in infrastructure-less scenarios. In this work, we apply wireless sensor networks devices to this problem that has hitherto received little attention. We analyze three points of the solution space. At one extreme, group membership information is proactively and collectively maintained by each node in the group. At the other extreme, the dissemination of group membership updates is triggered reactively by relying on a lower-level neighbor discovery protocol. In the middle lies a solution borrowing ideas from the two extremes. We compare our solutions through simulation of synthetic scenarios and real-world mobility traces of humans.