TWPalo: Through-the-wall passive localization of moving human with Wi-Fi

Abstract

Recently, we have witnessed a booming growth of Wi-Fi-based indoor localization systems that can locate the user without requiring the user to hold or wear any device. To foster a broad range of real-world based applications via these systems, the localization under the through-the-wall scenario (TTW) needs to be addressed, which is of great challenge. Based on the in-depth understanding of human induced reflection conveyed by Wi-Fi channel state information (CSI), we propose TWPalo, a through-the-wall device-free localization system. To achieve this goal, a three-dimensional joint parameter estimation algorithm is proposed to estimate the time of flight (ToF), angle of arrival (AoA), and Doppler frequency shift (DFS) of the propagation path. Then, by iterating joint parameter estimation, channel reconstruction, and cancellation, an innovative algorithm is proposed to separate the CSI of each propagation path and obtain the parameters of subtle human induced reflection (HIR), which could be overwhelmed by strong propagation paths, under the TTW scenario. At last, TWPalo translates the obtained AoA and ToF of subtle HIR into the locations of the human behind the wall. Extensive experimental evaluation validates the great performance of TWPalo in terms of parameter estimation and localization, under typical indoor TTW scenarios of glass and concrete wall.