Secure and Real-Time Behavior Prediction of Elderly Data Using RSA-AES Hybrid Encryption and ANN with Backpropagation Method

The security of private information in elderly mobile monitoring applications is basic to secure client security and guarantee reliable Caregiver. This think about proposes a hybrid encryption system that combines RSA and AES-256 calculations to protect both enrollment and real-time action information of elderly clients. An ID trade component is used to safely connect elderly clients with their assigned Caregivers. In addition, each login session triggers the era of a new encryption key, improving security against inactive key assaults. Real-time action information, counting development status and wellbeing measurements, is encrypted powerfully and put away in an upgraded “current” record structure, with chosen areas cleared out decoded for productive questioning. A decrypted component empowers authorized Caregivers to get to the elderly user's movement information safely and consistently. The system also integrates to an Artificial Neural Network (ANN) with the Backpropagation method to predict elderly behavior in real time. Sensor data from the accelerometer and gyroscope is used as input for the ANN, which classifies in vary of activities. The backpropagation algorithm optimizes the model by minimizing prediction errors, enabling it to learn complex sensor patterns and support early detection of abnormal behaviors. The results demonstrate that the proposed system effectively maintains data confidentiality, integrity, and access control, while also providing accurate real-time behavior analysis for elderly care applications.