A Received Signal Strength Based Fall Detection System Using Cognitive Sensing

Activity detection using radio signals has attracted a lot of research interest lately. A useful smart-healthcare application lies in monitoring an elderly person and anticipating possibly alarming situations. This paper proposes an elderly fall detection using radio frequency with cognitive sensing. The proposed approach uses cognitive sensing to identify the least used Wi-Fi channel and subsequently switches to that particular channel for least disturbances. The proposed method uses Wi-Fi (IEEE 802.11) to detect a person’s activity in a non-invasive manner. The system uses Received Signal Strength to trace the movements of the person. Using machine learning techniques, patterns associated with the person’s activities are identified. The experiments performed to demonstrate that the system can detect a human fall with an accuracy of around 70%.