Title:
Non-Contact Human Vital Sign Sensing and Activities Recognition with Radar and WiFi Comunications: Research Overview from Toda Laboratory in Nihon University

Abstract:
In this work, we present recent research activities conducted in the Toda Laboratory at Nihon University, focusing on non-contact human vital sign (heart rate and blood pressure) sensing and activity recognition using MIMO mmWave radar and Wi-Fi communication, exploring practical applications and commercialization opportunities. For human activity recognition, we introduce the classification of nine postures using a MIMO radar and a MIMO Wi-Fi network. Furthermore, we present a method for estimating 18 skeletal joint positions using a 12Tx–16Rx imaging radar. For hardware solutions, we use a low-power solution using a 2Tx–4Rx MIMO 60 GHz-band radar equipped with a single in-phase (I-only) mixer, comparing with the conventional 79 GHz-band 3Tx–4Rx MIMO radar employing both in-phase and quadrature (I/Q) mixers. Additionally, we explore a further lower power consumption 1Tx–1Rx 60 GHz-band pulsed radar. On the other hand, for vital sign sensing, we utilize channel state information (CSI) obtained from 3Tx–3Rx MIMO Wi-Fi routers. Signal processing techniques include beamforming for multi-target tracking and heart rate extraction using traditional frequency-time domain analysis. On the other hand, deep learning-based methods are employed for the human activity recognition and the skeletal pose estimation.