Device-free object tracking with wireless sensors

The device-free (DF) object tracking in wireless networks requires a tracked object neither be equipped with a transceiver nor actively participate in the tracking process. One of methods to implement a DF tracking system is to leverage the variation of radio signals introduced by object intrusion. However, many factors may result in the variation of radio signals and it is hard to distinguish which variation is caused by object intrusion. In this paper, we investigate how some factors impact signal variations, and in further impact the effectiveness of DF tracking. Based on analysis, we conduct experiments with TelosB motes, which are carried out in three representative environments with different system parameters. Different wireless channel quality indicators are employed to monitor signal variations. The experiment results provide a significant guidance on how to choose environments, system parameters, channel quality indicators, and metrics to implement an effective DF tracking system.