A self-calibrated centroid localization algorithm for indoor ZigBee WSNs

Abstract

A self-calibrated centroid localization algorithm is presented for wireless sensor networks based on IEEE 802.15.4 radio standard. Beside common issues in ranging accuracy using either received signal strength (RSSI) or link quality (LQI), it is challenging to cope with the fast changing environments. Although good accuracy has been achieved with methods such as Weighted Maximum Likelihood Estimation (WMLE) and Weighted Centroid Localization (WCL), these methods assume that environmental parameters, such as antenna gain, humidity, temperature, metallic surfaces, radio interference and other objects in the neighborhood remain constant over the localization period. To improve accuracy in a practical system where environment is continuously changing, the Self-Calibrating Centroid Localization (SCCL) algorithm is introduced in this paper. Simulation shows that SCCL can reduce the localization error by 65% as compared to WCL.