Using Environmental Light and Wireless Signals for Enhanced Mobility Relationship Analysis

With the rise of smartphones and wearables, research into human movement, particularly group interactions, has gained significant attention. These devices serve as invaluable sources of diverse data streams, encompassing movement patterns, social interactions, and historical wireless signals crucial for group identification. However, the accuracy of wireless signal recognition could be better in close proximity scenarios. Our paper presents a pioneering methodology to overcome this limitation that integrates wireless signal data with environmental light sensor features. This novel fusion enhances mobility group recognition accuracy significantly. We conducted rigorous experiments in a building. Our method consistently demonstrated robust performance in identifying mobility groups accurately. This empirical validation underscores the effectiveness and adaptability of our proposed approach.