Music guided SVD method for the detection of vital signs with SFCW radar

Abstract The detection of vital signs in complex environments using radar has a wide range of applications, from search and rescue to military purposes. Life detection can be achieved by detecting vital signs such as heartbeat, breathing, and movements of body parts. Because of the complex nature of the rubble environments, radar data contain clutter and noise that makes it difficult to detect signals with low amplitude levels, especially respiratory movements arised from breathing. In this study, data were collected with Stepped-Frequency Continuous Waveform (SFCW) radar for five scenarios in different environments. The method presented was used to determine whether living beings were in the environment. The data containing only vital signs was reconstructed using Singular Value Decomposition (SVD), and the range profile was obtained using the Multiple Signal Classification (MUSIC) algorithm. The range profiles of both the raw data and the data reconstructed with SVD were used to make the detections. The detections on the obtained range profile were later re-examined using Singular Spectrum Analysis to further reduce the false alarm rate. As a result of this process, data taken from different environments have shown that a decision can be made as to whether there is a vital sign in the environment.