The intelligent monitoring for the elderly based on WiFi signals

Abstract

With the increasing aging population, the intelligent monitoring for the elderly living alone has become a hot research topic. As the universal of WiFi, the intelligent monitoring system based on WiFi Channel State Information (CSI) is proposed for elderly, including fall detection and sleep monitoring. The system is completely off-the-shelf and non-contact, which is more convenient and comfortable to users. The fall detection algorithm is based on phase difference and amplitude of CSI, whose features of different motions are extracted, and the falls are detected by machine learning algorithm. The sleep monitoring system only used phase difference changes of antennas; meanwhile, the respiration, turning over, and apnea will be detected, which can be used to evaluate sleep quality as well. According to the results of experiments, the accuracy of the system can reach 84.6% in fall detection. In sleep monitoring, the error of each breath is 0.337 s and the average accuracy is 89.9%.