Radio Based Device Activity Recognition

Abstract

Recognizing human activities in their daily living allows the event and wide usage of human-centric applications, like health observance, aided living, etc. ancient activity recognition ways typically believe physical sensors (camera, measuring device, gyroscope, etc.) to endlessly collect sensing element readings, and utilize pattern recognition algorithms to spot user's activities at an aggregator. Though ancient activity recognition ways are incontestable to be effective in previous work, they raise some issues like privacy, energy consumption and preparation value. In recent years, a brand new activity recognition approach, that takes advantage of body attenuation and/or channel weakening of wireless radio, has been planned. Compared with ancient activity recognition ways, radio primarily based ways utilize wireless transceivers in environments as infrastructure, exploit radio communication characters to attain high recognition accuracy, scale back energy value and preserve user's privacy. During this paper, we tend to divide radio ways into four categories: ZigBee radio based activity recognition, local area network radio primarily based activity recognition, RFID radio primarily based activity recognition, and different radio primarily based activity recognition. Some existing add every class is introduced and reviewed thoroughly. Then, we tend to compare some representative ways to point out their blessings and downsides. At last, we tend to entails some future analysis directions of this new analysis topic.