Implementation and testing of a secure fall detection system for Body Area Networks

As the average age of the people in the Western world increases, so does the number of orthopaedic medical treatments due to falls. A method for better treatment of this type of injuries consists of early and accurate fall detection. We present an implementation and test of the system for accurate fall detection. First, a number of tests were performed to get thresholds for distinguishing falls from fall-like or normal behaviour. Then, the algorithm was implemented and tested on Biomobius, a technology platform, which allows researcher to rapidly develop sophisticated technology solutions for biomedical research. Sensors for gathering the acceleration data were part of the Tyndall Mote, a wireless sensor platform designed at Tyndall National Institute, Cork. Data is transmitted wirelessly to the PC. Therefore, since security has to be also considered in such medical systems, an encryption system is also implemented and tested on a low power microprocessor.