Dynamic clustering for tracking multiple transceiver-free objects

RF-based transceiver-free object tracking, originally proposed by the authors, allows real-time tracking of a moving object, where the object does not have to be equipped with an RF transceiver. Our previous algorithm, the best cover algorithm, suffers from a drawback, i.e., it does not work well when there are multiple objects in the tracking area. In this paper, we propose a localization model of distance, transmission power and the signal dynamics caused by the objects. The signal dynamics are derived from the measured Radio Signal Strength Indication (RSSI). Using this new model, we propose the "probabilistic cover algorithm" which is based on distributed dynamic clustering thus it can dramatically improve the localization accuracy when multiple objects are present. Moreover, the probabilistic cover algorithm can reduce the tracking latency in the system. We argue that the small overhead of the proposed algorithm makes it scalable for large deployment. Experimental results show that in addition to its ability to identify multiple objects, the tracking accuracy is improved at a rate of 10% to 20%.