Gesture Detection Using IEEE 802.15.4z-Compliant Impulse Radio UWB Radar System

This paper presents an IEEE 802.15.4z-compliant impulse radio ultra-wideband (IR-UWB) radar system for gesture detection. The presented IR-UWB radar system detects whether one of 12 predefined gestures has been performed by a person sitting on the passenger seat of a vehicle. The defined gestures include various hand movements, e.g., swiping left-right, or rotating the hand clock-wise. For the gesture detection, a neural networks (NNs)-based classifier is employed. The designed NNs requires less than 20 KByte of memory to store the parameters, which makes it suitable for embedded implementation for in-cabin sensing applications within the automotive industry. The designed NN achieves an average accuracy of more than 99% for the measurements data, showing its potential for enhanced interaction between the passenger and the vehicle's control systems.