The detection of human motions (e.g., walking, running, falling, etc.) without any physical contact with the human subjects plays an important role in a number of scientific and engineering applications such as biomedical diagnosis and treatment, physical condition monitoring of athletes in training process, health monitoring for senior citizens in long-term care facilities, urban operations for military and homeland security, and remote monitoring of the astronauts living status in orbital spacecrafts. Numerous radar technologies have been used for this purpose; such as Continuous-wave (CW) Doppler radar, Ultra-wideband (UWB) radars and stepped-frequency (SF) CW radars for experimental investigation of their applicability. However, pure experimental approaches lack precise control over the subject's motions, and thus can't effectively isolate the contributions from different components of the human motion for a good assessment. Electromagnetic modeling of human motion in a realistic fashion and accurately is thus important. In this work, I present such a human model capable of realistic human motions, and utilize full-wave computational electromagnetics techniques on hardware accelerated platforms to accurately predict the radar response for such a large scale problem. To the best of my knowledge, modeling of human motion for radar detection has not been studied. Thus, this research will focus on modeling human motion and its remote detection with radar technology in indoor environments to aid optimizing and enhancing experimental approaches. The proposed research consists of three main components: (1) development of a realistic human motion model, (2) development of an accurate electromagnetic scattering model from the scene, and (3) development of a detection and motion classification algorithm.