A Simple and Fast Human Activity Recognition System Using Radio Frequency Energy Harvesting

Abstract

Recognizing indoor activities of an individual provides useful information in smart living, well-being monitoring, and fitness management. In this paper, we propose a simple and fast human activity recognition (HAR) system based on Radio Frequency energy harvesting (RFEH). The intuition is that the harvested voltage signals of different human activities exhibit distinctive patterns. Utilizing the data collected from four smartphones, the RFEH-based HAR system indicates over 91% accuracy of activity recognition across all devices. By combining the lightweight classifiers and making an ensemble classification, an overall accuracy of over 97% is achieved.