Contents I. Introduction
A variety of wireless vital sign detection sensors have been studied for non-contact long term monitoring of patient conditions in past decades. In most cases continuous wave Doppler radars are utilized for remote detection systems, and their accuracy and reliabilities are demonstrated [1]–[4]. These systems can detect vital signs (breathing and heartbeat) from distances of 1 to 2 m, however, they have still many issues to be overcome. One of the main hurdles is the random movement of human subject itself. Doppler sensors reveal high sensitivity for detection of a tiny heartbeat signal, but large signals are also produced by random body movement. The heartbeat signal is often buried by other motion signals, such as shaking heads, swing arms, etc. To resolve the motion effects, dual-radar systems are proposed and certain degrees of random body motion have been successfully eliminated [5], [6]. The other detection problem arises when there are multiple targets. Single continuous wave (CW) Doppler radar presents difficulties in detecting separate vital signs from multiple targets without using phased array [3] or multiple antenna sensors [7]. The most recent works on detection of random movement of a subject have utilized pulsed Doppler or FMCW radars, and video cameras [8]–[10], and a tag attached to a subject, which serves as a sensor beacon, has also been introduced [11]. Thus, it is clear that motion artifacts and multiple subject problems are a significant burden in radar type vital sign detection systems.
