Enabling Mutual Surveillance in Wireless Sensor Networks

Keeping a group of persons or items geographically close together can be challenging task. In this report we consider the continuous self-surveillance of wireless sensor network (WSN) tags attached to individual group members. In such a herding system the tags check whether a given group is “together”, i.e. whether the group members are close to each other. We introduce our specific framework to the design of future herding system in which the sensor nodes transmit regular beacons, and on this basis cooperatively check for the presence or absence of individual group members. To enable cooperation the tags are required to form monitoring groups to reliably monitor each group member over time. Especially in case of dense networks an individual tag is not able to monitor all of its neighbors but has to restrict to a subset. We present and evaluate a truly distributed approach to the problem of forming such monitoring groups called Distributed Randomized Selection (DRS). This approach is well able to adapt to different network densities, ensures that all tags are monitored by a sufficient number of neighbors and requires no extra control packets besides the regular beacons.