This paper proposes a fall detection approach to detect accidental falls for senior citizens using wireless sensor networks based health-care systems.Unlike traditional sensory-context based detection

this paper achieves this goal by using radio signals in wireless sensor networks.A group of the sensors on the senior citizens’body (wearable sensors) and several sensors as anchor nodes (anchor sensors) in their daily territories are deployed.The fall event is detected by comparing the difference of signal strengths from wearable sensors caused by the natural characteristics of the wireless radio signals attenuation through human bodies.Both centralized and distributed lightweight algorithms are proposed to compute the status of the fall event from senior citizens’normal daily behaviors.The obtained evaluation and experiment results using MicaZ motes show that this approach performs high accuracy

low overhead and less response delay in fall events detection and report.