Different human activities (walking, falling, sitting, waving) cause distinct disturbances in the Wi-Fi signal multipath. By feeding CSI amplitude and phase data into Machine Learning (ML) classifiers (such as SVMs or Deep Neural Networks), systems built on the Atheros CSI Tool can recognize activities without cameras (device-free sensing).
The is a robust, open-source platform designed for fine-grained wireless channel measurements using commercial off-the-shelf (COTS) Atheros 802.11n Wi-Fi hardware. Unlike standard tools that only provide basic connection metrics, this tool extracts Channel State Information (CSI) , offering a high-resolution "image" of how signals interact with their environment at the physical (PHY) layer. Core Capabilities and Technical Features atheros csi tool