Device-Free Human Activity Recognition in Through-the-Wall Scenarios Using Single-Link Wi-Fi Channel Measurements

Device-free sensing using Wi-Fi channel state information (CSI) is a nonintrusive method of human activity recognition (HAR) in an urban built environment. In this article, we propose an HAR model for through-the-wall scenarios based on a novel feature set and machine learning models. The proposed feature set leverages time- and frequency-domain characteristics of single-link Wi-Fi channel measurements, collected from a low-cost Internet of Things (IoT) device. Diverse nonline-of-sight (LoS) conditions are analyzed such as open and closed doors, occupants in multiple rooms, and transmitter-receiver spacing of up to 9 m, to constitute seven datasets. The performance of various machine learning models on the proposed feature set was compared, and high prediction accuracies between 97% and 99% were obtained on linear discriminant analysis (LDA) and tree-based ensemble learning models. Joint activity recognition and location classification were also examined and by combining the proposed time- and frequency-domain features, a mean cross-validation accuracy of 99.6% was obtained on the LDA classifier.