A Millimeter-Wave Scattering Channel Model for Indoor Human Activity Sensing

In sixth-generation (6G) application scenarios like industry 5.0, augmented reality (AR), autonomous transportation, and eHealth, there is a growing demand for Human Activity Recognition (HAR). Meanwhile, with the deployment of millimeter-wave (mmWave) technologies in fifth-generation (5G) cellular communications, higher-resolution sensing becomes feasible. Utilizing mmWave for communication and HAR has garnered attention, necessitating accurate modeling of sensing channels. This paper proposes a mmWave scattering channel model for indoor HAR, which facilitates system design, optimization, and implementation. In the proposed model, we integrate primitive-based human body scattering where the human body is indicated by a set of primitives, and cluster-based environment scattering models, enabling detailed modeling of self-shadowing and double-bounce environment scattering. Additionally, we develop a simulation framework encompassing signal transmission, sensing channels, and processing, allowing adjustment of system parameters. Simulation results indicated by micro-Doppler signatures including multi-link effects show good agreements with measurements, validating the effectiveness of the proposed model. Meanwhile, the time consumption of the proposed simulation workflow for generating micro-Doppler signatures for most human activities is within 10 minutes.