Study On An Effective Fall Assistance System With Non-repudiation

In the big trend of China’s growing aging of population, to offer timely and efficient rescue to the elderly is both of high economic value and social significance. However, due to the adverse effect of some historical cases, it is now often the case that an elderly falls to the ground with no one to help. Public controversy about “to help or not help when an elderly falls” is often heard, and the Ministry of Health also issued a related technical guidelines. But these are only guidelines from the aspect of moral and emergency methods, there is still an urgent need for some feasible technology-based methods. For this reason, this paper makes efforts on the study of a non-Repudiation fall assistance system, the main functions of which are to detect human falling, automatically call for help and perform the assistance process, and provide non-repudiation proof for active rescuer. The system is based on Android mobile intelligent terminals and there are two main points of innovation. A fall detection algorithm is designed in consideration of the limited energy and computing resources of mobile intelligent terminals. The algorithm collects acceleration data via the built-in tri-axis acceleration sensor, analyzes and extracts the user’s personalized falling characteristics through an adaptive machine learning method, and then makes real-time detection of falling. The fall detection algorithm has the characteristics of low complexity, low power consumption, high sensitivity and relatively high accuracy. Besides, in order to deal with the bad phenomenon that people frame the rescuer for hurting after they have been rescued from falling lying on the ground, a non-repudiation fall assistance scheme is proposed. In the scheme, the moment of having detected a falling event, the system will automatically broadcast a message with identity-based signature to users nearby for help. The content of the message includes basic personal information, patient history and first aid of the user who fell down. Users who receive the helping message get proper permission to read the information according to their choice of whether to help or not, and then perform as rescuers or witnesses with the information they’ve got. The whole process has security features including anti-repudiation, user privacy protection, anti-replay, anti-key-leaking attacks and so on. The entire system is designed to provide reliable proofs for people who offer help, and eliminate their worries about the result of enthusiastically helping. Not only can the system serve the elderly, but also be used as a convenient application to provide reliable health care for users of all ages.