Gait-based Authentication in Smart Aging Care Systems

A Smart Aging Care System (SACS) allows elderly individuals to age gracefully in their homes through a connected network of heterogeneous sensors, devices, and third-party applications. However, like any Internet of Things (IoT) system, a SACS is susceptible to security threats.Gait-based authentication is a non-invasive and cost-effective biometric modality that enhances user experience (e.g., automatic authentication of a smart pill dispenser) and security. However, gait patterns become more pronounced with age, making it challenging to recognize elderly individuals. Given the limited emphasis on gait identification for the elderly, this work proposes an enhanced gait identification technique.The results show that we could identify 96.20% of gait data pertaining to the correct individual due to properly accounting for intra-subject gait fluctuations. We also propose an intrusion detection system to safeguard gait data against data manipulation attacks, that considers human factors in flagging a behavioral event as anomalous or normal.