Optimal design of millimeter wave MIMO radar for gesture sensing

Abstract

Gesture sensing is a favorable functionality of smart devices nowadays to allow users to interact with machines with hand gestures. Among different frequency bands, millimeter wave (mmWave) radar is most suitable for gesture sensing because it can achieve high range resolution in short range. Besides, an emerging technology, multiple-input multiple-output (MIMO), can help increase angular resolution with a saved number of antenna elements. In this project, we used MATLAB to simulate a mmWave MIMO radar system. The simulation includes transmitting and receiving antenna arrays, target position, transmitted and received signal, and radar signal processing such as matched filtering and digital beamforming. The range and angular resolution were compared for different radar configurations under different signal-to-noise ratios (SNR) and different target settings. The benefits of MIMO radar that it can increase angular resolution with fewer antenna elements used were illustrated in simulation experiments. At the end of the report, we proposed an optimal radar design for gesture sensing, which can balance the functionality of radar and physical space consumption. The transmitted signal we used was a 60.5 GHz FMCW signal, which is commonly used in commercial radars. We utilized two 3TX 4RX antenna arrays to measure three hand waving movements in front-back, left-right, and up-down directions for simple and reliable gesture sensing in harsh environment. Simulation experiments have shown that the radar design can recognize and differentiate three basic movements: front-back, left-right, and up-down movement with high precision.