Tracking Objects in a Smart Home

Abstract

A RSSI-based localization system on a home wireless sensor network is proposed in this work. In order to support a robot assistant in pick-and-place tasks, our current system is capable of estimating the localization of an object using the signal strength received by a mobile device in a ZigBee sensor network. Two models were utilized (a) log-distance path loss - model in which signal lost has a random influence with log-normal distribution, and (b) free space decay law - based on the decay law for a signal on an open space. RSSI measurements were done in laboratory for applying the estimation method. Moreover experiments with satisfactory results were done with a public dataset to benchmark our results.