The acquisition of vital signs used to be exclusively a clinical procedure requiring specialized equipment such as a cardiograph. Moreover, the medical examination itself, traditionally, requires instruments attached to the patient. Typical examples include, but are not limited to, the electrocardiogram (ECG) or the electroencephalogram (EEG). While these contact-based monitoring systems are considered the gold standard, they present major limitations: they are often invasive, uncomfortable, and may interfere with critical procedures or patient mobility. For example, for patients with psychiatric conditions, adherence to wearable sensors is not effective, undermining data reliability. To address these challenges, contactless monitoring systems have been proposed with the goal to remotely sense the desired vital signs. Even though different techniques can be used for acquiring different types of vital signs, those linked to body movement can be effectively detected using radar principles. Under this light, a millimeter wave (mmW) radar system is a very good candidate as it enables detection of small movements on close range due to its small wavelength. Radar-based solutions operating in the mmW spectrum have been proposed for estimating respiration and heart rates (Prat et al., 2019, doi: 10.1109/TAP.2018.2889595). While these measurements provide a useful starting point, the acquired signals remain far from clinical gold standards and even farther from the electrocardiogram (ECG), whose rich information is closely linked to various patient conditions. To bridge this gap, the Doppler Cardiograph (DCG) was introduced, aiming to correlate radar-based heartbeat responses with the heart's underlying electrical activity (Dong et al., 2020, doi: 10.1109/TMTT.2019.2948844). A fundamental challenge, however, lies in the fact that cardiac motion traces are strongly contaminated by lung and torso movements, as well as by voluntary and involuntary muscular activity, making it extremely difficult to isolate the heart's motion. Despite numerous investigations, no robust solution has yet emerged to enable accurate, remotely sensed ECG. Furthermore, although research interest in this area is growing, most existing systems remain proprietary and confined to in-house implementations. This lack of accessible platforms limits interdisciplinary collaboration; for instance, computer scientists seeking to apply advanced signal-processing or machine-learning methods are often unable to do so without first building dedicated radar hardware. In this work, we address these challenges and propose practical solutions. Specifically, we introduce a radar-based approach that exploits a wide frequency range and demonstrates promising results for remote ECG sensing. In addition, we present a cost-effective laboratory platform built entirely with commercial off-the-shelf (COTS) components and open-source GNU Radio software. The proposed system is presented in Fig. 1 and it consists of (a) the X410 USRP software defined radio (SDR) for the baseband signal, (b) the HMC6350 RX/TX device which upconverts and downconverts the baseband signal to 57-64 GHz, and (c) two WR-15 rectangular horn antennas.
