Using Wireless Channel Fading for Human Activity Recognition

This paper presents a way how to recognize of human activity by considering the fading characteristics of received signal strength (RSS) at the frequency 2.4 GHz on the standard of IEEE 802.11. The research use dominant attribute two things such as average of RSS and fluctuation analysis (FA) to cover efficiency of the presentation measure RSS from the implementation of four activities by using the supervised learning in order to spread the pattern of recognition from the test data obtained from the experiment three locations for the distance between a transmitter and a receiver at 5 meters for line-of-sight (LOS) case, this approach can provide accuracy not less than 90% [1]. On the average in the canal Obstructed-Line-of-sight (OLOS) give the lower accuracy compared to the case LOS but still give more than 80% in the case of test canal by implementation changing people found that the average result is 88.75%, in the case use another AP the average result is 90.63% and also in the case test NLOS. Finally, from exploratory experiments the proposed technique shows as a potential scheme to be adopted in activity-based recognition for many future applications like intrusion detection for home security, elderly personal surveillance, etc.