"Wearable-Free Multi-Node Indoor Gait Monitoring Dataset for Dementia Behaviour Analysis"

"This dataset presents a wearable-free indoor gait monitoring system designed to capture movement behaviours relevant to dementia assessment and smart healthcare research. The dataset is collected using a distributed multi-node sensing architecture consisting of four synchronized sensing nodes deployed in an indoor environment. Each node continuously records distance-based time-series measurements that reflect subject movement dynamics across different spatial regions. The dataset includes multiple participants performing dementia-related gait patterns such as direct walking, pacing, lapping, and random wandering. These behavioural patterns are commonly associated with cognitive decline and mobility irregularities observed in elderly populations. To ensure uniformity and enable long-duration behavioural modelling, all activity sessions are standardized to fixed temporal lengths through pattern-preserving time-series extension while maintaining original gait dynamics and multi-node correlation characteristics. The dataset provides synchronized spatial sensing information suitable for research in ambient intelligence, non-intrusive activity recognition, indoor localization, IoT-based healthcare monitoring, and edge artificial intelligence. By offering a multi-sensor, wearable-free sensing perspective, the dataset supports the development and benchmarking of intelligent monitoring systems aimed at early detection and continuous assessment of dementia-related behavioural changes.  "