Experimental Study: A LQI-Based Ranging Technique in ZigBee Sensor Networks

Abstract

Ranging technology which estimates the distance between two communicating wireless nodes has been widely used as a necessary component in localization solutions for wireless sensor networks (WSN). LQI (Link Quality Indicator) is a metric introduced in IEEE 802.15.4 that measures the error in the incoming modulation of successfully received packets that pass the CRC (Cyclic Redundancy Check). Because of low system cost and less computational complexity, LQI-based ranging techniques are increasingly applied in Zigbee sensor networks. However, due to the environmental affects and electronic noise generated by hardware, raw LQI data could not be directly aligned with distances. To eliminate errors in LQI data and obtain higher ranging accuracy, we design and evaluate a novel LQI-based ranging technique which includes three essential data processing components: pre-correction, error compensation and mixed regression analysis. First, anchor nodes with known locations are used in pre-correction process to correct LQI measurements against the empirical regression function obtained from historical data. Then, error compensation is applied to eliminate the intrinsic error in LQI data. Finally, ranging results are refined by the mixed regression analysis. The proposed ranging technique is implemented and evaluated on a Zigbee sensor prototype Tarax. Experiment results show that the average ranging error is less than 1m, confirming that the proposed technique is able to achieve higher ranging accuracy and suitable for localization applications in WSN.