Equivalent Pipeline Processing for IR-UWB and FMCW Radar Comparison in Vital Signs Monitoring Applications

Abstract

Several studies have been showing radars as promising contactless vital signs monitoring tools, with IR-UWB and FMCW, being the most popular radar architectures implemented for vital signs remote sensing applications. However, most of the studies focus on developing algorithms for processing data extracted from one only type of radar and consequently, neglecting the effect that hardware choices have on their results and algorithms performance. To carry out a more complete pilot study and deal with the algorithms and hardware influences, we use two off-the-shelf radars with IR-UWB and FMCW characteristics and make them agnostic through the implementation of a data equivalent processing pipeline. To test it, a pilot test was developed from where the vital signs of 10 healthy participants were recorded with both radars and compared with “ground truth” data. Up to four radar-algorithms pairs were compared where an absolute minimum error of 0.3 respiration per minute for respiratory rate and 1.8 beats per minute for heartbeat rate were obtained among the four configurations. From our results, IR-UWB architecture was seen as the best option to be implemented in a bio-radar application. We recommend our equivalent processing pipeline method to assess bio-radar systems in a more realistic way.