A Real-Time System for Wireless Human Activity Sensing and Indoor Localization

Abstract

Indoor localization and human activity sensing have been studied in many perspectives in recent years. Researchers are making efforts to achieve better indoor localization accuracy, adaption ability and less cost. Many contributions are made for indoor localization, but most of them either require dedicated infrastructures like RF radar, Bluetooth, acoustic beacons, or unique fingerprints for different environments. Human activity related works are using RF radars, COTS Wi-Fi devices, acoustic sensors and cameras as the data source. The latest researches show wireless signals are used to sense the movements of a human, even recognize gestures, gaits, heart rate, respiration rate, and rough body figure of a human. However, few of these works are widely used, and few of the works can be applied in real time. In this paper, a real-time system is introduced to sense the movements of moving humans in the room, and at the same time, locate all connecting clients with an accuracy of 2 meters. This system only requires commercial off-the-shelf Network Interface Cards. Because of the limitation of commercial off-the-shelf Network Interface Cards, the real-time localization is not fully implemented in the real-time system, but it can be used separately. It can be used in homes for security, in buildings to depict the occupancy map of rooms, or in shops to prevent unprivileged customers getting access to the free Wi-Fi.