Implementation of a 24 GHz Radar Combined with Accelerometers for Detection of Passengers in a Moving Vehicle

Abstract

In applying a continuous-wave Doppler radar for detection of passengers inside a vehicle, it has been questioned whether the vehicle’s movement would interfere on the radar’s detection capability. Once the vehicle and radar are moving, not only the passenger’s vital signs are detected, but also noise signals caused by engine vibrations, road bumps and relative movements between the vehicle and the passenger. This thesis approaches this question joining a radar and an accelerometer into one single device for supplementary measurement of the vehicle’s acceleration and, therefore, its movement. With the accelerometer information about the vehicle’s motion, the radar signal can later be filtered to find a cleaner signal of the passenger’s vital signs. Additionally, accelerometer belts were crafted and attached to the body of car passengers to measure their acceleration and movement as well. A complete software with user interface for acquiring and processing radar and accelerometer signals was developed for this project and tests have been performed in real road environments. Initial analyses of the collected data have shown that it is indeed possible to detect the vital signs of passengers despite the noise caused by the car’s motion. Accordingly, this thesis develops the software and integrates the hardware into a real vehicle system, paving the way for further investigations on the relation between the movements of a vehicle and radar detection.