Development of a device-free fall detection system in home environment

Abstract

In this report, a simple methodology was proposed to identify falling entity using a wireless sensor network located in an indoor environment. Most of the other methodologies such as installing a camera and wearing an electronics devices are being proposed but this method does not required any personnel to carry any devices and at the same time detecting the human falls without invading any privacies. For this methodology, the monitored environment will have eight sensor nodes and each different parallel layer and height will be deployed four sensor nodes to monitor the received signal strength indicator (RSSI). The sensor nodes are call Bluetooth Low Energy (BLE) which will be programmed using Waspmote Application Programming Interface (API) function. The BLE will be working in pairs, either one will be a Master or a Slave to collect the changes in RSSI data that used to detect the presences of human and human falls. The data will be concluded through a falling algorithm using Kalman filter and a Median approach. Different network orientation and several test results will be executed to prove the effectiveness of this methodology.