**1. Introduction**

Research on vital-signs detection using radar technology has been performed since the short-range radar system was introduced for the detection of human vital signs in the 1970s [1]. A vital-signs detector based on radar technology can measure the vital signs without electrode contacts or restriction of the measurement environment, in contrast to contact-type detectors [2]. Owing to these advantages, the radar sensor for vital-signs detection is promising as a key element in continuous monitoring systems for home-care service, local positioning and tracking in disaster scenes such as earthquakes and fires, as well as disease identification using heartrate variability (HRV) analysis, e.g., sleep apnea and angina pectoris [3–5]. In particular, the study on a vital-signs detector using radar technology aims to achieve a level of detection accuracy so as to fully replace the electrocardiogram (ECG) sensor in medical applications, such as HRV analysis.

Among the radar technologies, continuous-wave (CW) Doppler radars are useful for vital-signs monitoring based on periodic motions in the human body because of their simple hardware configuration and signal processing [3–6]. However, CW radars have a limitation in that noise can be increased or the receiver can be saturated by the movement of the target or surrounding clutters, as signals caused by all movements are collected by the antenna [7]. This limitation causes particularly significant degradation of accuracy in heartbeat detection compared with respiration detection, because the chest movement caused by the heartbeat is 0.2–0.5 mm, whereas the chest movement caused by respiration is 4–12 mm [8,9]. It is important to implement the CW radar with a high signal-to-noise ratio (SNR) for heartbeat detection to increase the detection accuracy [10]. The SNR in the CW radar can be improved by increasing the transmitted power, but the maximum allowable effective isotropic radiated power is restricted by the regulation in each frequency band. It is not easy to implement a radar front-end with a high SNR by using a low-noise and high-gain design methodology, and a complex radar architecture including calibration circuits and a calibration process could be needed to improve the SNR [10]. The accuracy of vital-signs detection can be also improved by using signal-processing techniques such as autocorrelation, the wavelet transform, and cross-correlation [11–14]. The autocorrelation method improves the detection accuracy for the periodic signal by converging signals to the most representative period frequency but has limitations for accurately monitoring heartbeat signals that change over time and evaluating their variability [11,14]. The wavelet transform is useful for increasing the accuracy by effectively extracting peaks of the heartbeat signal, but it is difficult to develop a generalized wavelet function that can improve the accuracy while being independent of the measurement environment, such as the characteristics of the subject and the clutter in the surroundings [12,14]. The cross-correlation method was proposed for increasing the detection accuracy by exploiting the similarity between vital signs independently measured at multiple frequencies by using a dual-band antenna and commercialized measurement equipment [13]. Although this approach is useful for achieving a high accuracy because the vital signs are independent of the characteristics of the radar and the operation frequency, it is necessary to use several radars with different operating frequencies, and customized components such as a dual-band antenna are required.

In this work, a vital-signs detector with improved accuracy based on frequency-shift keying (FSK) radar technology is proposed. An FSK radar is used for a highly accurate range detection based on the phase difference between the transmitting and receiving signals separately obtained at more than two operating frequencies [15]. The vital-signs detection in the FSK radar is performed by using the same method as the CW Doppler radar as the FSK radar can be regarded as operating with several CW signals in the same hardware configuration. The proposed FSK radar improves the detection accuracy and SNR by using the cross-correlation between the vital signs independently obtained from each operating frequency. A method of effectively discriminating two frequency signals at the baseband output is necessary for accurately obtaining the results of the cross-correlation in the FSK radar [16]. This work presents the signal discrimination technique based on envelope detection in synchronization with a frequency control signal as a method for separately obtaining each operating frequency signal. The proposed technique can discriminate the phase information at each operating frequency from the FSK radar in a short period, and the cross-correlation for improving the detection accuracy can be easily implemented in the single radar by using this technique. The imbalance between the in-phase (*I*) and quadrature (*Q*) signals and a direct-current (DC) offset, which are critical characteristics for the accuracy of the phase measurement, are calibrated by modifying the method used by the CW Doppler radar sensor for the FSK radar. Distance measurements with the proposed radar indicate that the phase difference of vital signs at each operating frequency have a cross-correlation. The measurement results of heartbeat detection have high accuracy relative to the reference electrocardiogram (ECG) signal owing to the cross-correlation of the proposed FSK radar. The measurement results for the distance and heartbeat show that the detection accuracy at each frequency depends on the distance, owing to the characteristics of the FSK radar. The operating principles of the FSK radar for vital-signs detection are described in Section 2. Section 3 shows the implementation method of the proposed FSK radar, including the calibration process and digital signal processing. The measurement results for vital signs and distances and an analysis of the results discussed in Section 4. Conclusions are presented in Section 5.
