This study investigates the influence of step length variations on the distance–time characteristics of human walking using a 24 GHz FMCW radar in a non-contact manner, and demonstrates the feasibility of estimating a subject’s height based on the correlation among step length, pulse count, and round-trip repetitions. Radar echo signals acquired from the head, chest, and lower-body regions were comparatively analyzed, revealing that the chest region provides the highest signal stability and repeatability, and thus serves as the optimal reference area for gait analysis. To enhance signal continuity and robustness, the acquired data were processed using windowing, interpolation, and outlier removal techniques, effectively mitigating noise-induced discontinuities. Two experimental scenarios were considered. In Case 1, the number of round trips was fixed while the step length was varied, and the required number of pulses was analyzed. The results indicate that an increase in step length leads to a reduction in the number of pulses required to traverse the same distance. In Case 2, the pulse count was fixed and the number of round trips was observed as a function of step length, showing that larger step lengths result in a greater number of round-trip motions within the same pulse duration. Based on the experimental data from both cases, an exponential regression model was derived, enabling continuous prediction of unmeasured step lengths or round-trip conditions. Furthermore, the interrelationship among pulse count, step length, and round-trip repetitions demonstrates the potential for estimating a subject’s height. This work presents a novel non-contact analysis framework that enables integrated estimation of step length, walking patterns, round-trip repetitions, and height using a single 24 GHz radar system, providing a foundational basis for applications in indoor gait monitoring, human behavior analysis, smart environments, and healthcare systems.
Kim et al. (Thu,) studied this question.