Implementation of a Flexible Real-Time Radar Architecture of Vital Signs Based on Cyclic Temporal Moment Algorithms Using SDR Technology

This article presents a low-cost bistatic radar built using software-defined radio (SDR) for real-time monitoring and detection of mechanical vibrations and physiological movements of a person’s chest. The proposed prototype is based on the complex received signal, which contains information about the heartbeat and respiration rates (RRs). Various algorithms can be employed for this purpose, and the choice depends on factors, such as accuracy, computational efficiency, and the application’s specific requirements. This study aims to evaluate and validate our signal processing solutions based on second-order cyclostationarity algorithms to distinguish cycle frequencies, using a low-cost universal software radio peripheral (USRP) SDR board. This device provides a software-defined RF architecture to prototype a vital signs radar with custom signal processing. The design concept for breathing and heart rates (HRs) was implemented and evaluated on a 5.8-GHz SDR platform (USRP2901) using GNU Radio combined with Python blocks. To verify the detecting performances of the proposed prototype, a series of simulations and experiments were conducted on different persons under tests to validate the theory. Moreover, two types of antennas were used: 1) an industrial horn antenna and 2) a patch antenna developed in our laboratory. The results were compared to a reference system, a pulse oximeter, to validate the performances of the algorithm. The respiration and heartbeat rates were correctly estimated for all the subjects with both antennas.