WIPP: Wi-Fi Compass for Indoor Passive Positioning with Decimeter Accuracy

Abstract

In recent decades, the proliferation of smart phones, tablets, and wireless networks has fostered a growing interest in indoor passive positioning. The Wi-Fi-based passive positioning systems can provide Location-Based Services (LBSs) to the third party such as market and security departments. Most of the existing systems are based on the Receive Signal Strength Indication (RSSI) information, which are generally time-consuming and susceptible to environmental change. To overcome this problem, we propose the Wi-Fi compass for indoor passive positioning system (WIPP), an Angle of Arrival (AOA) based passive positioning system using the existing commodity Wi-Fi network. In this paper, we first propose a new algorithm for the joint estimation of AOA and Time of Arrival (TOA) measurements based on the fine-grained Channel State Information (CSI), which is collected by an off-the-shelf Wi-Fi device equipped with only three antennas. Second, we use the affinity propagation clustering algorithm to identify the direct signal path from the target to each Wi-Fi Access Point (AP). Finally, we deploy the WIPP in an actual indoor environment to conduct the performance comparison with the well-known radio-frequency (RF) based system for locating and tracking users inside buildings (RADAR), as well as the conventional passive positioning system using the AOA solely. The experimental results show that the WIPP is able to achieve the median positioning error 0.7 m, which is much lower than the ones by the RADAR system and the conventional system using the AOA solely.