Research on HAR-Based Floor Positioning

Floor positioning is an important aspect of indoor technology, which closely related to location-based services (LBSs). Currently, floor technologies are mainly based on radio signals and barometric pressure. The former impacted by the multipath effect, rely infrastructure support, limited different spatial structures. For latter, air pressure changes with temperature humidity, deployment cost reference station high, terminal models need be calibrated in advance. In view these issues, here, we propose a novel method human activity recognition (HAR), using smartphone built-in sensor data classify pedestrian activities. We obtain degree change according category every step determine whether completes switching through condition threshold analysis. Then, combine previous or high-precision initial calculate pedestrians’ real-time position. A multi-floor office building was chosen as experimental site verified process alternating multiple types results show that position location accuracy this were high 100%, has good robustness universality. It more stable than methods wireless signals. Compared one existing HAR-based pressure, paper allows pedestrians undertake long-term static round-trip activities during going up down stairs. addition, proposed fault tolerance for misjudgment actions.